Remote Sensing

14 pages, 4648 KB

Open AccessEditor’s ChoiceLetter

U-Net-Id, an Instance Segmentation Model for Building Extraction from Satellite Images—Case Study in the Joanópolis City, Brazil

by Fabien H. Wagner, Ricardo Dalagnol, Yuliya Tarabalka, Tassiana Y. F. Segantine, Rogério Thomé and Mayumi C. M. Hirye

Remote Sens. 2020, 12(10), 1544; https://doi.org/10.3390/rs12101544 - 12 May 2020

Cited by 59 | Viewed by 14389

Abstract

Currently, there exists a growing demand for individual building mapping in regions of rapid urban growth in less-developed countries. Most existing methods can segment buildings but cannot discriminate adjacent buildings. Here, we present a new convolutional neural network architecture (CNN) called U-net-id that [...] Read more.

Currently, there exists a growing demand for individual building mapping in regions of rapid urban growth in less-developed countries. Most existing methods can segment buildings but cannot discriminate adjacent buildings. Here, we present a new convolutional neural network architecture (CNN) called U-net-id that performs building instance segmentation. The proposed network is trained with WorldView-3 satellite RGB images (0.3 m) and three different labeled masks. The first is the building mask; the second is the border mask, which is the border of the building segment with 4 pixels added outside and 3 pixels inside; and the third is the inner segment mask, which is the segment of the building diminished by 2 pixels. The architecture consists of three parallel paths, one for each mask, all starting with a U-net model. To accurately capture the overlap between the masks, all activation layers of the U-nets are copied and concatenated on each path and sent to two additional convolutional layers before the output activation layers. The method was tested with a dataset of 7563 manually delineated individual buildings of the city of Joanópolis-SP, Brazil. On this dataset, the semantic segmentation showed an overall accuracy of 97.67% and an F1-Score of 0.937 and the building individual instance segmentation showed good performance with a mean intersection over union (IoU) of 0.582 (median IoU = 0.694). Full article

(This article belongs to the Special Issue Data Mining and Machine Learning in Urban Applications)

► Show Figures

Graphical abstract

20 pages, 2125 KB

Open AccessFeature PaperEditor’s ChoiceArticle

Mapping Floristic Patterns of Trees in Peruvian Amazonia Using Remote Sensing and Machine Learning

by Pablo Pérez Chaves, Gabriela Zuquim, Kalle Ruokolainen, Jasper Van doninck, Risto Kalliola, Elvira Gómez Rivero and Hanna Tuomisto

Remote Sens. 2020, 12(9), 1523; https://doi.org/10.3390/rs12091523 - 10 May 2020

Cited by 9 | Viewed by 10386

Abstract

Recognition of the spatial variation in tree species composition is a necessary precondition for wise management and conservation of forests. In the Peruvian Amazonia, this goal is not yet achieved mostly because adequate species inventory data has been lacking. The recently started Peruvian [...] Read more.

Recognition of the spatial variation in tree species composition is a necessary precondition for wise management and conservation of forests. In the Peruvian Amazonia, this goal is not yet achieved mostly because adequate species inventory data has been lacking. The recently started Peruvian national forest inventory (INFFS) is expected to change the situation. Here, we analyzed genus-level variation, summarized through non-metric multidimensional scaling (NMDS), in a set of 157 INFFS inventory plots in lowland to low mountain rain forests (<2000 m above sea level) using Landsat satellite imagery and climatic, edaphic, and elevation data as predictor variables. Genus-level floristic patterns have earlier been found to be indicative of species-level patterns. In correlation tests, the floristic variation of tree genera was most strongly related to Landsat variables and secondly to climatic variables. We used random forest regression, under varying criteria of feature selection and cross-validation, to predict the floristic composition on the basis of Landsat and environmental data. The best model explained >60% of the variation along NMDS axes 1 and 2 and 40% of the variation along NMDS axis 3. We used this model to predict the three NMDS dimensions at a 450-m resolution over all of the Peruvian Amazonia and classified the pixels into 10 floristic classes using k-means classification. An indicator analysis identified statistically significant indicator genera for 8 out of the 10 classes. The results are congruent with earlier studies, suggesting that the approach is robust and can be applied to other tropical regions, which is useful for reducing research gaps and for identifying suitable areas for conservation. Full article

(This article belongs to the Special Issue Remote Sensing for Biodiversity Mapping and Monitoring)

► Show Figures

Graphical abstract

22 pages, 3399 KB

Open AccessEditor’s ChoiceReview

Deep Learning Approaches Applied to Remote Sensing Datasets for Road Extraction: A State-Of-The-Art Review

by Arnick Abdollahi, Biswajeet Pradhan, Nagesh Shukla, Subrata Chakraborty and Abdullah Alamri

Remote Sens. 2020, 12(9), 1444; https://doi.org/10.3390/rs12091444 - 2 May 2020

Cited by 254 | Viewed by 17508

Abstract

One of the most challenging research subjects in remote sensing is feature extraction, such as road features, from remote sensing images. Such an extraction influences multiple scenes, including map updating, traffic management, emergency tasks, road monitoring, and others. Therefore, a systematic review of [...] Read more.

One of the most challenging research subjects in remote sensing is feature extraction, such as road features, from remote sensing images. Such an extraction influences multiple scenes, including map updating, traffic management, emergency tasks, road monitoring, and others. Therefore, a systematic review of deep learning techniques applied to common remote sensing benchmarks for road extraction is conducted in this study. The research is conducted based on four main types of deep learning methods, namely, the GANs model, deconvolutional networks, FCNs, and patch-based CNNs models. We also compare these various deep learning models applied to remote sensing datasets to show which method performs well in extracting road parts from high-resolution remote sensing images. Moreover, we describe future research directions and research gaps. Results indicate that the largest reported performance record is related to the deconvolutional nets applied to remote sensing images, and the F1 score metric of the generative adversarial network model, DenseNet method, and FCN-32 applied to UAV and Google Earth images are high: 96.08%, 95.72%, and 94.59%, respectively. Full article

(This article belongs to the Special Issue Traffic Assessment and Monitoring with Remote Sensing and Geospatial Modelling)

► Show Figures

Graphical abstract

19 pages, 10695 KB

Open AccessEditor’s ChoiceArticle

Integrating National Ecological Observatory Network (NEON) Airborne Remote Sensing and In-Situ Data for Optimal Tree Species Classification

by Victoria M. Scholl, Megan E. Cattau, Maxwell B. Joseph and Jennifer K. Balch

Remote Sens. 2020, 12(9), 1414; https://doi.org/10.3390/rs12091414 - 30 Apr 2020

Cited by 41 | Viewed by 10363

Abstract

Accurately mapping tree species composition and diversity is a critical step towards spatially explicit and species-specific ecological understanding. The National Ecological Observatory Network (NEON) is a valuable source of open ecological data across the United States. Freely available NEON data include in-situ measurements [...] Read more.

Accurately mapping tree species composition and diversity is a critical step towards spatially explicit and species-specific ecological understanding. The National Ecological Observatory Network (NEON) is a valuable source of open ecological data across the United States. Freely available NEON data include in-situ measurements of individual trees, including stem locations, species, and crown diameter, along with the NEON Airborne Observation Platform (AOP) airborne remote sensing imagery, including hyperspectral, multispectral, and light detection and ranging (LiDAR) data products. An important aspect of predicting species using remote sensing data is creating high-quality training sets for optimal classification purposes. Ultimately, manually creating training data is an expensive and time-consuming task that relies on human analyst decisions and may require external data sets or information. We combine in-situ and airborne remote sensing NEON data to evaluate the impact of automated training set preparation and a novel data preprocessing workflow on classifying the four dominant subalpine coniferous tree species at the Niwot Ridge Mountain Research Station forested NEON site in Colorado, USA. We trained pixel-based Random Forest (RF) machine learning models using a series of training data sets along with remote sensing raster data as descriptive features. The highest classification accuracies, 69% and 60% based on internal RF error assessment and an independent validation set, respectively, were obtained using circular tree crown polygons created with half the maximum crown diameter per tree. LiDAR-derived data products were the most important features for species classification, followed by vegetation indices. This work contributes to the open development of well-labeled training data sets for forest composition mapping using openly available NEON data without requiring external data collection, manual delineation steps, or site-specific parameters. Full article

(This article belongs to the Special Issue She Maps)

► Show Figures

Graphical abstract

14 pages, 1563 KB

Open AccessEditor’s ChoiceArticle

Compatibility of Aerial and Terrestrial LiDAR for Quantifying Forest Structural Diversity

by Elizabeth A. LaRue, Franklin W. Wagner, Songlin Fei, Jeff W. Atkins, Robert T. Fahey, Christopher M. Gough and Brady S. Hardiman

Remote Sens. 2020, 12(9), 1407; https://doi.org/10.3390/rs12091407 - 29 Apr 2020

Cited by 67 | Viewed by 9477

Abstract

Structural diversity is a key feature of forest ecosystems that influences ecosystem functions from local to macroscales. The ability to measure structural diversity in forests with varying ecological composition and management history can improve the understanding of linkages between forest structure and ecosystem [...] Read more.

Structural diversity is a key feature of forest ecosystems that influences ecosystem functions from local to macroscales. The ability to measure structural diversity in forests with varying ecological composition and management history can improve the understanding of linkages between forest structure and ecosystem functioning. Terrestrial LiDAR has often been used to provide a detailed characterization of structural diversity at local scales, but it is largely unknown whether these same structural features are detectable using aerial LiDAR data that are available across larger spatial scales. We used univariate and multivariate analyses to quantify cross-compatibility of structural diversity metrics from terrestrial versus aerial LiDAR in seven National Ecological Observatory Network sites across the eastern USA. We found strong univariate agreement between terrestrial and aerial LiDAR metrics of canopy height, openness, internal heterogeneity, and leaf area, but found marginal agreement between metrics that described heterogeneity of the outermost layer of the canopy. Terrestrial and aerial LiDAR both demonstrated the ability to distinguish forest sites from structural diversity metrics in multivariate space, but terrestrial LiDAR was able to resolve finer-scale detail within sites. Our findings indicated that aerial LiDAR could be of use in quantifying broad-scale variation in structural diversity across macroscales. Full article

(This article belongs to the Special Issue Lidar Remote Sensing of Forest Structure, Biomass and Dynamics)

► Show Figures

Graphical abstract

24 pages, 3122 KB

Open AccessEditor’s ChoiceArticle

High Quality Zenith Tropospheric Delay Estimation Using a Low-Cost Dual-Frequency Receiver and Relative Antenna Calibration

by Andreas Krietemeyer, Hans van der Marel, Nick van de Giesen and Marie-Claire ten Veldhuis

Remote Sens. 2020, 12(9), 1393; https://doi.org/10.3390/rs12091393 - 28 Apr 2020

Cited by 42 | Viewed by 7049

Abstract

The recent release of consumer-grade dual-frequency receivers sparked scientific interest into use of these cost-efficient devices for high precision positioning and tropospheric delay estimations. Previous analyses with low-cost single-frequency receivers showed promising results for the estimation of Zenith Tropospheric Delays (ZTDs). However, their [...] Read more.

The recent release of consumer-grade dual-frequency receivers sparked scientific interest into use of these cost-efficient devices for high precision positioning and tropospheric delay estimations. Previous analyses with low-cost single-frequency receivers showed promising results for the estimation of Zenith Tropospheric Delays (ZTDs). However, their application is limited by the need to account for the ionospheric delay. In this paper we investigate the potential of a low-cost dual-frequency receiver (U-blox ZED-F9P) in combination with a range of different quality antennas. We show that the receiver itself is very well capable of achieving high-quality ZTD estimations. The limiting factor is the quality of the receiving antenna. To improve the applicability of mass-market antennas, a relative antenna calibration is performed, and new absolute Antenna Exchange Format (ANTEX) entries are created using a geodetic antenna as base. The performance of ZTD estimation with the tested antennas is evaluated, with and without antenna Phase Center Variation (PCV) corrections, using Precise Point Positioning (PPP). Without applying PCVs for the low-cost antennas, the Root Mean Square Errors (RMSE) of the estimated ZTDs are between 15 mm and 24 mm. Using the newly generated PCVs, the RMSE is reduced significantly to about 4 mm, a level that is excellent for meteorological applications. The standard U-blox ANN-MB-00 patch antenna, with a circular ground plane, after correcting the phase pattern yields comparable results (0.47 mm bias and 4.02 mm RMSE) to those from geodetic quality antennas, providing an all-round low-cost solution. The relative antenna calibration method presented in this paper opens the way for wide-spread application of low-cost receiver and antennas. Full article

(This article belongs to the Special Issue Remote Sensing of Precipitation: Part II)

► Show Figures

Graphical abstract

20 pages, 8235 KB

Open AccessFeature PaperEditor’s ChoiceArticle

Near Real-Time Monitoring of the Christmas 2018 Etna Eruption Using SEVIRI and Products Validation

by Stefano Corradini, Lorenzo Guerrieri, Dario Stelitano, Giuseppe Salerno, Simona Scollo, Luca Merucci, Michele Prestifilippo, Massimo Musacchio, Malvina Silvestri, Valerio Lombardo and Tommaso Caltabiano

Remote Sens. 2020, 12(8), 1336; https://doi.org/10.3390/rs12081336 - 23 Apr 2020

Cited by 33 | Viewed by 4793

Abstract

On the morning of 24 December 2018, an eruptive event occurred at Etna, which was followed the next day by a strong sequence of shallow earthquakes. The eruptive episode lasted until 30 December, ranging from moderate strombolian to lava fountain activity coupled with [...] Read more.

On the morning of 24 December 2018, an eruptive event occurred at Etna, which was followed the next day by a strong sequence of shallow earthquakes. The eruptive episode lasted until 30 December, ranging from moderate strombolian to lava fountain activity coupled with vigorous ash/gas emissions and a lava flow effusion toward the eastern volcano flank of Valle del Bove. In this work, the data collected from the Spinning Enhanced Visible and InfraRed Imager (SEVIRI) instruments on board the Meteosat Second Generation (MSG) geostationary satellite are used to characterize the Etna activity by estimating the proximal and distal eruption parameters in near real time. The inversion of data indicates the onset of eruption on 24 December at 11:15 UTC, a maximum Time Average Discharge Rate (TADR) of 8.3 m³/s, a cumulative lava volume emitted of 0.5 Mm³, and a Volcanic Plume Top Height (VPTH) that reached a maximum altitude of 8 km above sea level (asl). The volcanic cloud ash and SO₂ result totally collocated, with an ash amount generally lower than SO₂ except on 24 December during the climax phase. A total amount of about 100 and 35 kt of SO₂ and ash respectively was emitted during the entire eruptive period, while the SO₂ fluxes reached peaks of more than 600 kg/s, with a mean value of about 185 kg/s. The SEVIRI VPTH, ash/SO₂ masses, and flux time series have been compared with the results obtained from the ground-based visible (VIS) cameras and FLux Automatic MEasurements (FLAME) networks, and the satellite images collected by the MODerate resolution Imaging Spectroradiometer (MODIS) instruments on board the Terra and Aqua- polar satellites. The analysis indicates good agreement between SEVIRI, VIS camera, and MODIS retrievals with VPTH, ash, and SO₂ estimations all within measurement errors. The SEVIRI and FLAME SO₂ flux retrievals show significant discrepancies due to the presence of volcanic ash and a gap of data on the FLAME network. The results obtained in this study show the ability of geostationary satellite systems to characterize eruptive events from the source to the atmosphere in near real time during the day and night, thus offering a powerful tool to mitigate volcanic risk on both local population and airspace and to give insight on volcanic processes. Full article

(This article belongs to the Special Issue Convective and Volcanic Clouds (CVC))

► Show Figures

Graphical abstract

23 pages, 3230 KB

Open AccessEditor’s ChoiceArticle

Sun-Angle Effects on Remote-Sensing Phenology Observed and Modelled Using Himawari-8

by Xuanlong Ma, Alfredo Huete, Ngoc Nguyen Tran, Jian Bi, Sicong Gao and Yelu Zeng

Remote Sens. 2020, 12(8), 1339; https://doi.org/10.3390/rs12081339 - 23 Apr 2020

Cited by 44 | Viewed by 11119

Abstract

Satellite remote sensing of vegetation at regional to global scales is undertaken at considerable variations in solar zenith angle (SZA) across space and time, yet the extent to which these SZA variations matter for the retrieval of phenology remains largely unknown. Here we [...] Read more.

Satellite remote sensing of vegetation at regional to global scales is undertaken at considerable variations in solar zenith angle (SZA) across space and time, yet the extent to which these SZA variations matter for the retrieval of phenology remains largely unknown. Here we examined the effect of seasonal and spatial variations in SZA on retrieving vegetation phenology from time series of the Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) across a study area in southeastern Australia encompassing forest, woodland, and grassland sites. The vegetation indices (VI) data span two years and are from the Advanced Himawari Imager (AHI), which is onboard the Japanese Himawari-8 geostationary satellite. The semi-empirical RossThick-LiSparse-Reciprocal (RTLSR) bidirectional reflectance distribution function (BRDF) model was inverted for each spectral band on a daily basis using 10-minute reflectances acquired by H-8 AHI at different sun-view geometries for each site. The inverted RTLSR model was then used to forward calculate surface reflectance at three constant SZAs (20°, 40°, 60°) and one seasonally varying SZA (local solar noon), all normalised to nadir view. Time series of NDVI and EVI adjusted to different SZAs at nadir view were then computed, from which phenological metrics such as start and end of growing season were retrieved. Results showed that NDVI sensitivity to SZA was on average nearly five times greater than EVI sensitivity. VI sensitivity to SZA also varied among sites (biome types) and phenological stages, with NDVI sensitivity being higher during the minimum greenness period than during the peak greenness period. Seasonal SZA variations altered the temporal profiles of both NDVI and EVI, with more pronounced differences in magnitude among NDVI time series normalised to different SZAs. When using VI time series that allowed SZA to vary at local solar noon, the uncertainties in estimating start, peak, end, and length of growing season introduced by local solar noon varying SZA VI time series, were 7.5, 3.7, 6.5, and 11.3 days for NDVI, and 10.4, 11.9, 6.5, and 8.4 days for EVI respectively, compared to VI time series normalised to a constant SZA. Furthermore, the stronger SZA dependency of NDVI compared with EVI, resulted in up to two times higher uncertainty in estimating annual integrated VI, a commonly used remote-sensing proxy for vegetation productivity. Since commonly used satellite products are not generally normalised to a constant sun-angle across space and time, future studies to assess the sun-angle effects on satellite applications in agriculture, ecology, environment, and carbon science are urgently needed. Measurements taken by new-generation geostationary (GEO) satellites offer an important opportunity to refine this assessment at finer temporal scales. In addition, studies are needed to evaluate the suitability of different BRDF models for normalising sun-angle across a broad spectrum of vegetation structure, phenological stages and geographic locations. Only through continuous investigations on how sun-angle variations affect spatiotemporal vegetation dynamics and what is the best strategy to deal with it, can we achieve a more quantitative remote sensing of true signals of vegetation change across the entire globe and through time. Full article

(This article belongs to the Special Issue Earth Monitoring from A New Generation of Geostationary Satellites)

► Show Figures

Graphical abstract

24 pages, 2097 KB

Open AccessEditor’s ChoiceArticle

Relation of Photochemical Reflectance Indices Based on Different Wavelengths to the Parameters of Light Reactions in Photosystems I and II in Pea Plants

by Ekaterina Sukhova and Vladimir Sukhov

Remote Sens. 2020, 12(8), 1312; https://doi.org/10.3390/rs12081312 - 21 Apr 2020

Cited by 32 | Viewed by 4688

Abstract

Measurement and analysis of the numerous reflectance indices of plants is an effective approach for the remote sensing of plant physiological processes in agriculture and ecological monitoring. A photochemical reflectance index (PRI) plays an important role in this kind of remote sensing because [...] Read more.

Measurement and analysis of the numerous reflectance indices of plants is an effective approach for the remote sensing of plant physiological processes in agriculture and ecological monitoring. A photochemical reflectance index (PRI) plays an important role in this kind of remote sensing because it can be related to early changes in photosynthetic processes under the action of stressors (excess light, changes in temperature, drought, etc.). In particular, we previously showed that light-induced changes in PRIs could be strongly related to the energy-dependent component of the non-photochemical quenching in photosystem II. The aim of the present work was to undertake comparative analysis of the efficiency of using light-induced changes in PRIs (ΔPRIs) based on different wavelengths for the estimation of the parameters of photosynthetic light reactions (including the parameters of photosystem I). Pea plants were used in the investigation; the photosynthetic parameters were measured using the pulse-amplitude-modulated (PAM) fluorometer Dual-PAM-100 and the intensities of the reflected light were measured using the spectrometer S100. The ΔPRIs were calculated as ΔPRI(band,570), where the band was 531 nm for the typical PRI and 515, 525, 535, 545, or 555 nm for modified PRIs; 570 nm was the reference wavelength for all PRIs. There were several important results: (1) ∆PRI(525,570), ∆PRI(531,570), ∆PRI(535,570), and ∆PRI(545,570) could be used for estimation of most of the photosynthetic parameters under light only or under dark only conditions. (2) The combination of dark and light conditions decreased the efficiency of ∆PRIs for the estimation of the photosynthetic parameters; ∆PRI(535,570) and ∆PRI(545,570) had maximal efficiency under these conditions. (3) ∆PRI(515,570) and ∆PRI(525,570) mainly included the slow-relaxing component of PRI; in contrast, ∆PRI(531,570), ∆PRI(535,570), ∆PRI(545,570), and ∆PRI(555,570) mainly included the fast-relaxing component of PRI. These components were probably caused by different mechanisms. Full article

(This article belongs to the Section Remote Sensing in Agriculture and Vegetation)

► Show Figures

Graphical abstract

20 pages, 15631 KB

Open AccessFeature PaperEditor’s ChoiceArticle

Harmonized Landsat 8 and Sentinel-2 Time Series Data to Detect Irrigated Areas: An Application in Southern Italy

by Salvatore Falanga Bolognesi, Edoardo Pasolli, Oscar Rosario Belfiore, Carlo De Michele and Guido D’Urso

Remote Sens. 2020, 12(8), 1275; https://doi.org/10.3390/rs12081275 - 17 Apr 2020

Cited by 24 | Viewed by 8448

Abstract

Lack of accurate and up-to-date data associated with irrigated areas and related irrigation amounts is hampering the full implementation and compliance of the Water Framework Directive (WFD). In this paper, we describe the framework that we developed and implemented within the DIANA project [...] Read more.

Lack of accurate and up-to-date data associated with irrigated areas and related irrigation amounts is hampering the full implementation and compliance of the Water Framework Directive (WFD). In this paper, we describe the framework that we developed and implemented within the DIANA project to map the actual extent of irrigated areas in the Campania region (Southern Italy) during the 2018 irrigation season. For this purpose, we considered 202 images from the Harmonized Landsat Sentinel-2 (HLS) products (57 images from Landsat 8 and 145 images from Sentinel-2). Such data were preprocessed in order to extract a multitemporal Normalized Difference Vegetation Index (NDVI) map, which was then smoothed through a gap-filling algorithm. We further integrated data coming from high-resolution (4 km) global satellite precipitation Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN)-Cloud Classification System (CCS) products. We collected an extensive ground truth in the field represented by 2992 data points coming from three main thematic classes: bare soil and rainfed (class 0), herbaceous (class 1), and tree crop (class 2). This information was exploited to generate irrigated area maps by adopting a machine learning classification approach. We compared six different types of classifiers through a cross-validation approach and found that, in general, random forests, support vector machines, and boosted decision trees exhibited the best performances in terms of classification accuracy and robustness to different tested scenarios. We found an overall accuracy close to 90% in discriminating among the three thematic classes, which highlighted promising capabilities in the detection of irrigated areas from HLS products. Full article

(This article belongs to the Section Environmental Remote Sensing)

► Show Figures

Figure 1

25 pages, 1451 KB

Open AccessEditor’s ChoiceReview

An Overview of Platforms for Big Earth Observation Data Management and Analysis

by Vitor C. F. Gomes, Gilberto R. Queiroz and Karine R. Ferreira

Remote Sens. 2020, 12(8), 1253; https://doi.org/10.3390/rs12081253 - 16 Apr 2020

Cited by 240 | Viewed by 22259

Abstract

In recent years, Earth observation (EO) satellites have generated big amounts of geospatial data that are freely available for society and researchers. This scenario brings challenges for traditional spatial data infrastructures (SDI) to properly store, process, disseminate and analyze these big data sets. [...] Read more.

In recent years, Earth observation (EO) satellites have generated big amounts of geospatial data that are freely available for society and researchers. This scenario brings challenges for traditional spatial data infrastructures (SDI) to properly store, process, disseminate and analyze these big data sets. To meet these demands, novel technologies have been proposed and developed, based on cloud computing and distributed systems, such as array database systems, MapReduce systems and web services to access and process big Earth observation data. Currently, these technologies have been integrated into cutting edge platforms in order to support a new generation of SDI for big Earth observation data. This paper presents an overview of seven platforms for big Earth observation data management and analysis—Google Earth Engine (GEE), Sentinel Hub, Open Data Cube (ODC), System for Earth Observation Data Access, Processing and Analysis for Land Monitoring (SEPAL), openEO, JEODPP, and pipsCloud. We also provide a comparison of these platforms according to criteria that represent capabilities of the EO community interest. Full article

(This article belongs to the Special Issue Spatial Data Infrastructures for Big Geospatial Sensing Data)

► Show Figures

Graphical abstract

34 pages, 22244 KB

Open AccessEditor’s ChoiceArticle

The Status of Earth Observation Techniques in Monitoring High Mountain Environments at the Example of Pasterze Glacier, Austria: Data, Methods, Accuracies, Processes, and Scales

by Michael Avian, Christian Bauer, Matthias Schlögl, Barbara Widhalm, Karl-Heinz Gutjahr, Michael Paster, Christoph Hauer, Melina Frießenbichler, Anton Neureiter, Gernot Weyss, Peter Flödl, Gernot Seier and Wolfgang Sulzer

Remote Sens. 2020, 12(8), 1251; https://doi.org/10.3390/rs12081251 - 15 Apr 2020

Cited by 16 | Viewed by 8055

Abstract

Earth observation offers a variety of techniques for monitoring and characterizing geomorphic processes in high mountain environments. Terrestrial laserscanning and unmanned aerial vehicles provide very high resolution data with high accuracy. Automatic cameras have become a valuable source of information—mostly in a qualitative [...] Read more.

Earth observation offers a variety of techniques for monitoring and characterizing geomorphic processes in high mountain environments. Terrestrial laserscanning and unmanned aerial vehicles provide very high resolution data with high accuracy. Automatic cameras have become a valuable source of information—mostly in a qualitative manner—in recent years. The availability of satellite data with very high revisiting time has gained momentum through the European Space Agency’s Sentinel missions, offering new application potential for Earth observation. This paper reviews the status of recent techniques such as terrestrial laserscanning, remote sensed imagery, and synthetic aperture radar in monitoring high mountain environments with a particular focus on the impact of new platforms such as Sentinel-1 and -2 as well as unmanned aerial vehicles. The study area comprises the high mountain glacial environment at the Pasterze Glacier, Austria. The area is characterized by a highly dynamic geomorphological evolution and by being subject to intensive scientific research as well as long-term monitoring. We primarily evaluate landform classification and process characterization for: (i) the proglacial lake; (ii) icebergs; (iii) the glacier river; (iv) valley-bottom processes; (v) slope processes; and (vi) rock wall processes. We focus on assessing the potential of every single method both in spatial and temporal resolution in characterizing different geomorphic processes. Examples of the individual techniques are evaluated qualitatively and quantitatively in the context of: (i) morphometric analysis; (ii) applicability in high alpine regions; and (iii) comparability of the methods among themselves. The final frame of this article includes considerations on scale dependent process detectability and characterization potentials of these Earth observation methods, along with strengths and limitations in applying these methods in high alpine regions. Full article

(This article belongs to the Special Issue Imaging Floods and Glacier Geohazards with Remote Sensing)

► Show Figures

Graphical abstract

20 pages, 5178 KB

Open AccessEditor’s ChoiceArticle

Satellite Observations for Detecting and Forecasting Sea-Ice Conditions: A Summary of Advances Made in the SPICES Project by the EU’s Horizon 2020 Programme

by Marko Mäkynen, Jari Haapala, Giuseppe Aulicino, Beena Balan-Sarojini, Magdalena Balmaseda, Alexandru Gegiuc, Fanny Girard-Ardhuin, Stefan Hendricks, Georg Heygster, Larysa Istomina, Lars Kaleschke, Juha Karvonen, Thomas Krumpen, Mikko Lensu, Michael Mayer, Flavio Parmiggiani, Robert Ricker, Eero Rinne, Amelie Schmitt, Markku Similä, Steffen Tietsche, Rasmus Tonboe, Peter Wadhams, Mai Winstrup and Hao Zuo Show full author list Hide full author list

Remote Sens. 2020, 12(7), 1214; https://doi.org/10.3390/rs12071214 - 10 Apr 2020

Cited by 19 | Viewed by 8313

Abstract

The detection, monitoring, and forecasting of sea-ice conditions, including their extremes, is very important for ship navigation and offshore activities, and for monitoring of sea-ice processes and trends. We summarize here recent advances in the monitoring of sea-ice conditions and their extremes from [...] Read more.

The detection, monitoring, and forecasting of sea-ice conditions, including their extremes, is very important for ship navigation and offshore activities, and for monitoring of sea-ice processes and trends. We summarize here recent advances in the monitoring of sea-ice conditions and their extremes from satellite data as well as the development of sea-ice seasonal forecasting capabilities. Our results are the outcome of the three-year (2015–2018) SPICES (Space-borne Observations for Detecting and Forecasting Sea-Ice Cover Extremes) project funded by the EU’s Horizon 2020 programme. New SPICES sea-ice products include pancake ice thickness and degree of ice ridging based on synthetic aperture radar imagery, Arctic sea-ice volume and export derived from multisensor satellite data, and melt pond fraction and sea-ice concentration using Soil Moisture and Ocean Salinity (SMOS) radiometer data. Forecasts of July sea-ice conditions from initial conditions in May showed substantial improvement in some Arctic regions after adding sea-ice thickness (SIT) data to the model initialization. The SIT initialization also improved seasonal forecasts for years with extremely low summer sea-ice extent. New SPICES sea-ice products have a demonstrable level of maturity, and with a reasonable amount of further work they can be integrated into various operational sea-ice services. Full article

(This article belongs to the Special Issue Polar Sea Ice: Detection, Monitoring and Modeling)

► Show Figures

Figure 1

17 pages, 6852 KB

Open AccessEditor’s ChoiceArticle

by Disong Fu, Zijue Song, Xiaoling Zhang, Yunfei Wu, Minzheng Duan, Weiwei Pu, Zhiqiang Ma, Weijun Quan, Huaigang Zhou, Huizheng Che and Xiangao Xia

Remote Sens. 2020, 12(7), 1193; https://doi.org/10.3390/rs12071193 - 8 Apr 2020

Cited by 14 | Viewed by 3856

Abstract

Surface particulate matter with an aerodynamic diameter of <2.5 μm (PM_2.5) and column-integrated aerosol optical depth (AOD) exhibits substantial diurnal, daily, and yearly variabilities that are regionally dependent. The diversity of these temporal variabilities in urban and rural areas may imply [...] Read more.

Surface particulate matter with an aerodynamic diameter of <2.5 μm (PM_2.5) and column-integrated aerosol optical depth (AOD) exhibits substantial diurnal, daily, and yearly variabilities that are regionally dependent. The diversity of these temporal variabilities in urban and rural areas may imply the inherent mechanisms. A novel time-series analysis tool developed by Facebook, Prophet, is used to investigate the holiday, seasonal, and inter-annual patterns of PM_2.5 and AOD at a rural station (RU) and an urban station (UR) in Beijing. PM_2.5 shows a coherent decreasing tendency at both stations during 2014–2018, consistent with the implementation of the air pollution action plan at the end of 2013. RU is characterized by similar seasonal variations of AOD and PM_2.5, with the lowest values in winter and the highest in summer, which is opposite that at UR with maximum AOD, but minimum PM_2.5 in summer and minimum AOD, but maximum PM_2.5 in winter. During the National Day holiday (1–7 October), both AOD and PM_2.5 holiday components regularly shift from negative to positive departures, and the turning point generally occurs on October 4. AODs at both stations steadily increase throughout the daytime, which is most striking in winter. A morning rush hour peak of PM_2.5 (7:00–9:00 local standard time (LST)) and a second peak at night (23:00 LST) are observed at UR. PM_2.5 at RU often reaches minima (maxima) at around 12:00 LST (19:00 LST), about four hours later (earlier) than UR. The ratio of PM_2.5 to AOD (η) shows a decreasing tendency at both stations in the last four years, indicating a profound impact of the air quality control program. η at RU always begins to increase about 1–2 h earlier than that at UR during the daytime. Large spatial and temporal variations of η suggest that caution should be observed in the estimation of PM_2.5 from AOD. Full article

(This article belongs to the Special Issue Advances in Remote Sensing and Modeling of Fires and/or Tropospheric Composition in Asia)

► Show Figures

Figure 1

18 pages, 7241 KB

Open AccessEditor’s ChoiceArticle

Regional Dependence of Atmospheric Responses to Oceanic Eddies in the North Pacific Ocean

by Jinlin Ji, Jing Ma, Changming Dong, John C. H. Chiang and Dake Chen

Remote Sens. 2020, 12(7), 1161; https://doi.org/10.3390/rs12071161 - 4 Apr 2020

Cited by 13 | Viewed by 4327

Abstract

Based on sea surface height anomaly (SSHA) from satellite altimeter and microwave radiometer datasets, this study investigates atmospheric responses to oceanic eddies in four subdomains of the North Pacific Ocean with strongest eddy activity: Kuroshio Extension (KE), Subtropical Front (SF), California Coastal Current [...] Read more.

Based on sea surface height anomaly (SSHA) from satellite altimeter and microwave radiometer datasets, this study investigates atmospheric responses to oceanic eddies in four subdomains of the North Pacific Ocean with strongest eddy activity: Kuroshio Extension (KE), Subtropical Front (SF), California Coastal Current (CC) and Aleutian Islands (AI). Analyses show that anticyclonic eddies cause sea surface temperature, surface wind speed and precipitation rate to increase in all four subdomains, and vice versa. Through a further examination of the regional dependence of atmospheric responses to oceanic eddies, it is found that the strongest and the weakest surface wind speed responses (in winter and summer) are observed in the KE and AI region, respectively. For precipitation rate, seasonal variation of the atmospheric responses to oceanic eddies is strongest in winter and weakest in summer in the KE, CC and AI regions, but stronger in summer in the SF area. The reasons for such regional dependence and seasonality are the differences in the strength of SST anomalies, the vertical kinetic energy flux and atmospheric instability in the four subdomains. Full article

(This article belongs to the Section Ocean Remote Sensing)

► Show Figures

Graphical abstract

28 pages, 2087 KB

Open AccessFeature PaperEditor’s ChoiceReview

Remote Sensing of River Discharge: A Review and a Framing for the Discipline

by Colin J Gleason and Michael T Durand

Remote Sens. 2020, 12(7), 1107; https://doi.org/10.3390/rs12071107 - 31 Mar 2020

Cited by 136 | Viewed by 16323

Abstract

Remote sensing of river discharge (RSQ) is a burgeoning field rife with innovation. This innovation has resulted in a highly non-cohesive subfield of hydrology advancing at a rapid pace, and as a result misconceptions, mis-citations, and confusion are apparent among authors, readers, editors, [...] Read more.

Remote sensing of river discharge (RSQ) is a burgeoning field rife with innovation. This innovation has resulted in a highly non-cohesive subfield of hydrology advancing at a rapid pace, and as a result misconceptions, mis-citations, and confusion are apparent among authors, readers, editors, and reviewers. While the intellectually diverse subfield of RSQ practitioners can parse this confusion, the broader hydrology community views RSQ as a monolith and such confusion can be damaging. RSQ has not been comprehensively summarized over the past decade, and we believe that a summary of the recent literature has a potential to provide clarity to practitioners and general hydrologists alike. Therefore, we here summarize a broad swath of the literature, and find after our reading that the most appropriate way to summarize this literature is first by application area (into methods appropriate for gauged, semi-gauged, regionally gauged, politically ungauged, and totally ungauged basins) and next by methodology. We do not find categorizing by sensor useful, and everything from un-crewed aerial vehicles (UAVs) to satellites are considered here. Perhaps the most cogent theme to emerge from our reading is the need for context. All RSQ is employed in the service of furthering hydrologic understanding, and we argue that nearly all RSQ is useful in this pursuit provided it is properly contextualized. We argue that if authors place each new work into the correct application context, much confusion can be avoided, and we suggest a framework for such context here. Specifically, we define which RSQ techniques are and are not appropriate for ungauged basins, and further define what it means to be ‘ungauged’ in the context of RSQ. We also include political and economic realities of RSQ, as the objective of the field is sometimes to provide data purposefully cloistered by specific political decisions. This framing can enable RSQ to respond to hydrology at large with confidence and cohesion even in the face of methodological and application diversity evident within the literature. Finally, we embrace the intellectual diversity of RSQ and suggest the field is best served by a continuation of methodological proliferation rather than by a move toward orthodoxy and standardization. Full article

(This article belongs to the Special Issue Remote Sensing of Flow Velocity, Channel Bathymetry, and River Discharge)

► Show Figures

Figure 1

20 pages, 7725 KB

Open AccessEditor’s ChoiceArticle

Tree Species Classification of Drone Hyperspectral and RGB Imagery with Deep Learning Convolutional Neural Networks

by Somayeh Nezami, Ehsan Khoramshahi, Olli Nevalainen, Ilkka Pölönen and Eija Honkavaara

Remote Sens. 2020, 12(7), 1070; https://doi.org/10.3390/rs12071070 - 26 Mar 2020

Cited by 154 | Viewed by 15672

Abstract

Interest in drone solutions in forestry applications is growing. Using drones, datasets can be captured flexibly and at high spatial and temporal resolutions when needed. In forestry applications, fundamental tasks include the detection of individual trees, tree species classification, biomass estimation, etc. Deep [...] Read more.

Interest in drone solutions in forestry applications is growing. Using drones, datasets can be captured flexibly and at high spatial and temporal resolutions when needed. In forestry applications, fundamental tasks include the detection of individual trees, tree species classification, biomass estimation, etc. Deep neural networks (DNN) have shown superior results when comparing with conventional machine learning methods such as multi-layer perceptron (MLP) in cases of huge input data. The objective of this research is to investigate 3D convolutional neural networks (3D-CNN) to classify three major tree species in a boreal forest: pine, spruce, and birch. The proposed 3D-CNN models were employed to classify tree species in a test site in Finland. The classifiers were trained with a dataset of 3039 manually labelled trees. Then the accuracies were assessed by employing independent datasets of 803 records. To find the most efficient set of feature combination, we compare the performances of 3D-CNN models trained with hyperspectral (HS) channels, Red-Green-Blue (RGB) channels, and canopy height model (CHM), separately and combined. It is demonstrated that the proposed 3D-CNN model with RGB and HS layers produces the highest classification accuracy. The producer accuracy of the best 3D-CNN classifier on the test dataset were 99.6%, 94.8%, and 97.4% for pines, spruces, and birches, respectively. The best 3D-CNN classifier produced ~5% better classification accuracy than the MLP with all layers. Our results suggest that the proposed method provides excellent classification results with acceptable performance metrics for HS datasets. Our results show that pine class was detectable in most layers. Spruce was most detectable in RGB data, while birch was most detectable in the HS layers. Furthermore, the RGB datasets provide acceptable results for many low-accuracy applications. Full article

(This article belongs to the Special Issue Thematic Information Extraction and Application in Forests)

► Show Figures

Figure 1

39 pages, 12479 KB

Open AccessEditor’s ChoiceReview

Accounting for Training Data Error in Machine Learning Applied to Earth Observations

by Arthur Elmes, Hamed Alemohammad, Ryan Avery, Kelly Caylor, J. Ronald Eastman, Lewis Fishgold, Mark A. Friedl, Meha Jain, Divyani Kohli, Juan Carlos Laso Bayas, Dalton Lunga, Jessica L. McCarty, Robert Gilmore Pontius, Andrew B. Reinmann, John Rogan, Lei Song, Hristiana Stoynova, Su Ye, Zhuang-Fang Yi and Lyndon Estes

Remote Sens. 2020, 12(6), 1034; https://doi.org/10.3390/rs12061034 - 23 Mar 2020

Cited by 81 | Viewed by 16317

Abstract

Remote sensing, or Earth Observation (EO), is increasingly used to understand Earth system dynamics and create continuous and categorical maps of biophysical properties and land cover, especially based on recent advances in machine learning (ML). ML models typically require large, spatially explicit training [...] Read more.

Remote sensing, or Earth Observation (EO), is increasingly used to understand Earth system dynamics and create continuous and categorical maps of biophysical properties and land cover, especially based on recent advances in machine learning (ML). ML models typically require large, spatially explicit training datasets to make accurate predictions. Training data (TD) are typically generated by digitizing polygons on high spatial-resolution imagery, by collecting in situ data, or by using pre-existing datasets. TD are often assumed to accurately represent the truth, but in practice almost always have error, stemming from (1) sample design, and (2) sample collection errors. The latter is particularly relevant for image-interpreted TD, an increasingly commonly used method due to its practicality and the increasing training sample size requirements of modern ML algorithms. TD errors can cause substantial errors in the maps created using ML algorithms, which may impact map use and interpretation. Despite these potential errors and their real-world consequences for map-based decisions, TD error is often not accounted for or reported in EO research. Here we review the current practices for collecting and handling TD. We identify the sources of TD error, and illustrate their impacts using several case studies representing different EO applications (infrastructure mapping, global surface flux estimates, and agricultural monitoring), and provide guidelines for minimizing and accounting for TD errors. To harmonize terminology, we distinguish TD from three other classes of data that should be used to create and assess ML models: training reference data, used to assess the quality of TD during data generation; validation data, used to iteratively improve models; and map reference data, used only for final accuracy assessment. We focus primarily on TD, but our advice is generally applicable to all four classes, and we ground our review in established best practices for map accuracy assessment literature. EO researchers should start by determining the tolerable levels of map error and appropriate error metrics. Next, TD error should be minimized during sample design by choosing a representative spatio-temporal collection strategy, by using spatially and temporally relevant imagery and ancillary data sources during TD creation, and by selecting a set of legend definitions supported by the data. Furthermore, TD error can be minimized during the collection of individual samples by using consensus-based collection strategies, by directly comparing interpreted training observations against expert-generated training reference data to derive TD error metrics, and by providing image interpreters with thorough application-specific training. We strongly advise that TD error is incorporated in model outputs, either directly in bias and variance estimates or, at a minimum, by documenting the sources and implications of error. TD should be fully documented and made available via an open TD repository, allowing others to replicate and assess its use. To guide researchers in this process, we propose three tiers of TD error accounting standards. Finally, we advise researchers to clearly communicate the magnitude and impacts of TD error on map outputs, with specific consideration given to the likely map audience. Full article

(This article belongs to the Section Environmental Remote Sensing)

► Show Figures

Graphical abstract

25 pages, 1435 KB

Open AccessEditor’s ChoiceArticle

On the Performances of Trend and Change-Point Detection Methods for Remote Sensing Data

by Ana F. Militino, Mehdi Moradi and M. Dolores Ugarte

Remote Sens. 2020, 12(6), 1008; https://doi.org/10.3390/rs12061008 - 21 Mar 2020

Cited by 71 | Viewed by 11241

Abstract

Detecting change-points and trends are common tasks in the analysis of remote sensing data. Over the years, many different methods have been proposed for those purposes, including (modified) Mann–Kendall and Cox–Stuart tests for detecting trends; and Pettitt, Buishand range, Buishand U, standard normal [...] Read more.

Detecting change-points and trends are common tasks in the analysis of remote sensing data. Over the years, many different methods have been proposed for those purposes, including (modified) Mann–Kendall and Cox–Stuart tests for detecting trends; and Pettitt, Buishand range, Buishand U, standard normal homogeneity (Snh), Meanvar, structure change (Strucchange), breaks for additive season and trend (BFAST), and hierarchical divisive (E.divisive) for detecting change-points. In this paper, we describe a simulation study based on including different artificial, abrupt changes at different time-periods of image time series to assess the performances of such methods. The power of the test, type I error probability, and mean absolute error (MAE) were used as performance criteria, although MAE was only calculated for change-point detection methods. The study reveals that if the magnitude of change (or trend slope) is high, and/or the change does not occur in the first or last time-periods, the methods generally have a high power and a low MAE. However, in the presence of temporal autocorrelation, MAE raises, and the probability of introducing false positives increases noticeably. The modified versions of the Mann–Kendall method for autocorrelated data reduce/moderate its type I error probability, but this reduction comes with an important power diminution. In conclusion, taking a trade-off between the power of the test and type I error probability, we conclude that the original Mann–Kendall test is generally the preferable choice. Although Mann–Kendall is not able to identify the time-period of abrupt changes, it is more reliable than other methods when detecting the existence of such changes. Finally, we look for trend/change-points in land surface temperature (LST), day and night, via monthly MODIS images in Navarre, Spain, from January 2001 to December 2018. Full article

► Show Figures

Graphical abstract

40 pages, 2089 KB

Open AccessEditor’s ChoiceReview

Applications of Unmanned Aerial Vehicles in Cryosphere: Latest Advances and Prospects

by Clare Gaffey and Anshuman Bhardwaj

Remote Sens. 2020, 12(6), 948; https://doi.org/10.3390/rs12060948 - 15 Mar 2020

Cited by 129 | Viewed by 14526

Abstract

Owing to usual logistic hardships related to field-based cryospheric research, remote sensing has played a significant role in understanding the frozen components of the Earth system. Conventional spaceborne or airborne remote sensing platforms have their own merits and limitations. Unmanned aerial vehicles (UAVs) [...] Read more.

Owing to usual logistic hardships related to field-based cryospheric research, remote sensing has played a significant role in understanding the frozen components of the Earth system. Conventional spaceborne or airborne remote sensing platforms have their own merits and limitations. Unmanned aerial vehicles (UAVs) have emerged as a viable and inexpensive option for studying the cryospheric components at unprecedented spatiotemporal resolutions. UAVs are adaptable to various cryospheric research needs in terms of providing flexibility with data acquisition windows, revisits, data/sensor types (multispectral, hyperspectral, microwave, thermal/night imaging, Light Detection and Ranging (LiDAR), and photogrammetric stereos), viewing angles, flying altitudes, and overlap dimensions. Thus, UAVs have the potential to act as a bridging remote sensing platform between spatially discrete in situ observations and spatially continuous but coarser and costlier spaceborne or conventional airborne remote sensing. In recent years, a number of studies using UAVs for cryospheric research have been published. However, a holistic review discussing the methodological advancements, hardware and software improvements, results, and future prospects of such cryospheric studies is completely missing. In the present scenario of rapidly changing global and regional climate, studying cryospheric changes using UAVs is bound to gain further momentum and future studies will benefit from a balanced review on this topic. Our review covers the most recent applications of UAVs within glaciology, snow, permafrost, and polar research to support the continued development of high-resolution investigations of cryosphere. We also analyze the UAV and sensor hardware, and data acquisition and processing software in terms of popularity for cryospheric applications and revisit the existing UAV flying regulations in cold regions of the world. The recent usage of UAVs outlined in 103 case studies provide expertise that future investigators should base decisions on. Full article

(This article belongs to the Special Issue Applications of Remote Sensing in Glaciology)

► Show Figures

Figure 1

23 pages, 9520 KB

Open AccessEditor’s ChoiceArticle

Mapping Three Decades of Changes in the Brazilian Savanna Native Vegetation Using Landsat Data Processed in the Google Earth Engine Platform

by Ane Alencar, Julia Z. Shimbo, Felipe Lenti, Camila Balzani Marques, Bárbara Zimbres, Marcos Rosa, Vera Arruda, Isabel Castro, João Paulo Fernandes Márcico Ribeiro, Victória Varela, Isa Alencar, Valderli Piontekowski, Vivian Ribeiro, Mercedes M. C. Bustamante, Edson Eyji Sano and Mario Barroso

Remote Sens. 2020, 12(6), 924; https://doi.org/10.3390/rs12060924 - 13 Mar 2020

Cited by 182 | Viewed by 17243

Abstract

Widespread in the subtropics and tropics of the Southern Hemisphere, savannas are highly heterogeneous and seasonal natural vegetation types, which makes change detection (natural vs. anthropogenic) a challenging task. The Brazilian Cerrado represents the largest savanna in South America, and the most threatened [...] Read more.

Widespread in the subtropics and tropics of the Southern Hemisphere, savannas are highly heterogeneous and seasonal natural vegetation types, which makes change detection (natural vs. anthropogenic) a challenging task. The Brazilian Cerrado represents the largest savanna in South America, and the most threatened biome in Brazil owing to agricultural expansion. To assess the native Cerrado vegetation (NV) areas most susceptible to natural and anthropogenic change over time, we classified 33 years (1985–2017) of Landsat imagery available in the Google Earth Engine (GEE) platform. The classification strategy used combined empirical and statistical decision trees to generate reference maps for machine learning classification and a novel annual dataset of the predominant Cerrado NV types (forest, savanna, and grassland). We obtained annual NV maps with an average overall accuracy ranging from 87% (at level 1 NV classification) to 71% over the time series, distinguishing the three main NV types. This time series was then used to generate probability maps for each NV class. The native vegetation in the Cerrado biome declined at an average rate of 0.5% per year (748,687 ha yr⁻¹), mostly affecting forests and savannas. From 1985 to 2017, 24.7 million hectares of NV were lost, and now only 55% of the NV original distribution remains. Of the remnant NV in 2017 (112.6 million hectares), 65% has been stable over the years, while 12% changed among NV types, and 23% was converted to other land uses but is now in some level of secondary NV. Our results were fundamental in indicating areas with higher rates of change in a long time series in the Brazilian Cerrado and to highlight the challenges of mapping distinct NV types in a highly seasonal and heterogeneous savanna biome. Full article

(This article belongs to the Special Issue Remote Sensing of Savannas and Woodlands)

► Show Figures

Figure 1

25 pages, 14879 KB

Open AccessEditor’s ChoiceArticle

Combining InfraRed Thermography and UAV Digital Photogrammetry for the Protection and Conservation of Rupestrian Cultural Heritage Sites in Georgia: A Methodological Application

by William Frodella, Mikheil Elashvili, Daniele Spizzichino, Giovanni Gigli, Luka Adikashvili, Nikoloz Vacheishvili, Giorgi Kirkitadze, Akaki Nadaraia, Claudio Margottini and Nicola Casagli

Remote Sens. 2020, 12(5), 892; https://doi.org/10.3390/rs12050892 - 10 Mar 2020

Cited by 51 | Viewed by 7669

Abstract

The rock-cut city of Vardzia is an example of the extraordinary rupestrian cultural heritage of Georgia. The site, Byzantine in age, was carved in the steep tuff slopes of the Erusheti mountains, and due to its peculiar geological characteristics, it is particularly vulnerable [...] Read more.

The rock-cut city of Vardzia is an example of the extraordinary rupestrian cultural heritage of Georgia. The site, Byzantine in age, was carved in the steep tuff slopes of the Erusheti mountains, and due to its peculiar geological characteristics, it is particularly vulnerable to weathering and degradation, as well as frequent instability phenomena. These problems determine serious constraints on the future conservation of the site, as well as the safety of the visitors. This paper focuses on the implementation of a site-specific methodology, based on the integration of advanced remote sensing techniques, such as InfraRed Thermography (IRT) and Unmanned Aerial Vehicle (UAV)-based Digital Photogrammetry (DP), with traditional field surveys and laboratory analyses, with the aim of mapping the potential criticality of the rupestrian complex on a slope scale. The adopted methodology proved to be a useful tool for the detection of areas of weathering and degradation on the tuff cliffs, such as moisture and seepage sectors related to the ephemeral drainage network of the slope. These insights provided valuable support for the design and implementation of sustainable mitigation works, to be profitably used in the management plan of the site of Vardzia, and can be used for the protection and conservation of rupestrian cultural heritage sites characterized by similar geological contexts. Full article

(This article belongs to the Special Issue Remote Sensing, Spatial Analysis, and GIS for Natural and Cultural Heritage Documentation, Monitoring, and Preservation)

► Show Figures

Graphical abstract

22 pages, 30065 KB

Open AccessEditor’s ChoiceArticle

Sentinel-1 DInSAR for Monitoring Active Landslides in Critical Infrastructures: The Case of the Rules Reservoir (Southern Spain)

by Cristina Reyes-Carmona, Anna Barra, Jorge Pedro Galve, Oriol Monserrat, José Vicente Pérez-Peña, Rosa María Mateos, Davide Notti, Patricia Ruano, Agustín Millares, Juan López-Vinielles and José Miguel Azañón

Remote Sens. 2020, 12(5), 809; https://doi.org/10.3390/rs12050809 - 3 Mar 2020

Cited by 62 | Viewed by 9736

Abstract

Landslides in reservoir contexts are a well-recognised hazard that may lead to dangerous situations regarding infrastructures and people’s safety. Satellite-based radar interferometry is proving to be a reliable method to monitor the activity of landslides in such contexts. Here, we present a DInSAR [...] Read more.

Landslides in reservoir contexts are a well-recognised hazard that may lead to dangerous situations regarding infrastructures and people’s safety. Satellite-based radar interferometry is proving to be a reliable method to monitor the activity of landslides in such contexts. Here, we present a DInSAR (Differential Interferometric Synthetic Aperture Radar) analysis of Sentinel-1 images that exemplifies the usefulness of the technique to recognize and monitor landslides in the Rules Reservoir (Southern Spain). The integration of DInSAR results with a comprehensive geomorphological study allowed us to understand the typology, evolution and triggering factors of three active landslides: Lorenzo-1, Rules Viaduct and El Arrecife. We could distinguish between rotational and translational landslides and, thus, we evaluated the potential hazards related to these typologies, i.e., retrogression (Lorenzo-1 and Rules Viaduct landslides) or catastrophic slope failure (El Arrecife Landslide), respectively. We also observed how changes in the water level of the reservoir influence the landslide’s behaviour. Additionally, we were able to monitor the stability of the Rules Dam as well as detect the deformation of a highway viaduct that crosses a branch of the reservoir. Overall, we consider that other techniques must be applied to continue monitoring the movements, especially in the El Arrecife Landslide, in order to avoid future structural damages and fatalities. Full article

(This article belongs to the Special Issue Landslide Monitoring, Susceptibility, Hazard Assessment and Prediction with Remotely Sensed Big Data)

► Show Figures

Graphical abstract

17 pages, 2072 KB

Open AccessEditor’s ChoiceArticle

Using NDVI to Differentiate Wheat Genotypes Productivity Under Dryland and Irrigated Conditions

by Mohammed A. Naser, Raj Khosla, Louis Longchamps and Subash Dahal

Remote Sens. 2020, 12(5), 824; https://doi.org/10.3390/rs12050824 - 3 Mar 2020

Cited by 65 | Viewed by 7610

Abstract

Crop breeders are looking for tools to facilitate the screening of genotypes in field trials. Remote sensing-based indices such as normalized difference vegetative index (NDVI) are sensitive to biomass and nitrogen (N) variability in crop canopies. The objectives of this study were (i) [...] Read more.

Crop breeders are looking for tools to facilitate the screening of genotypes in field trials. Remote sensing-based indices such as normalized difference vegetative index (NDVI) are sensitive to biomass and nitrogen (N) variability in crop canopies. The objectives of this study were (i) to determine if proximal sensor-based NDVI readings can differentiate the yield of winter wheat (Triticum aestivum L.) genotypes and (ii) to determine if NDVI readings can be used to classify wheat genotypes into grain yield productivity classes. This study was conducted in northeastern Colorado in 2010 and 2011. The NDVI readings were acquired weekly from March to June, during 2010 and 2011. The correlation between NDVI and grain yield was determined using Pearson’s product-moment correlation coefficient (r). The k-means clustering method was used to classify mean NDVI and mean grain yield into three classes. The overall accuracy between NDVI and yield classes was reported. The findings of this study show that, under dryland conditions, there is a reliable correlation between grain yield and NDVI at the early growing season, at the anthesis growth stage, and the mid-grain filling growth stage, as well as a poor association under irrigated conditions. Our results suggest that when the sensor is not saturated, i.e., NDVI < 0.9, NDVI could assess grain yield with fair accuracy. This study demonstrated the potential of using NDVI readings as a tool to differentiate and identify superior wheat genotypes. Full article

(This article belongs to the Section Remote Sensing in Agriculture and Vegetation)

► Show Figures

Graphical abstract

31 pages, 14150 KB

Open AccessEditor’s ChoiceArticle

Land-Cover Changes to Surface-Water Buffers in the Midwestern USA: 25 Years of Landsat Data Analyses (1993–2017)

by Tedros M. Berhane, Charles R. Lane, Samson G. Mengistu, Jay Christensen, Heather E. Golden, Shi Qiu, Zhe Zhu and Qiusheng Wu

Remote Sens. 2020, 12(5), 754; https://doi.org/10.3390/rs12050754 - 25 Feb 2020

Cited by 18 | Viewed by 6156

Abstract

To understand the timing, extent, and magnitude of land use/land cover (LULC) change in buffer areas surrounding Midwestern US waters, we analyzed the full imagery archive (1982–2017) of three Landsat footprints covering ~100,000 km². The study area included urbanizing Chicago, Illinois [...] Read more.

To understand the timing, extent, and magnitude of land use/land cover (LULC) change in buffer areas surrounding Midwestern US waters, we analyzed the full imagery archive (1982–2017) of three Landsat footprints covering ~100,000 km². The study area included urbanizing Chicago, Illinois and St. Louis, Missouri regions and agriculturally dominated landscapes (i.e., Peoria, Illinois). The Continuous Change Detection and Classification algorithm identified 1993–2017 LULC change across three Landsat footprints and in 90 m buffers for ~110,000 surface waters; waters were also size-binned into five groups for buffer LULC change analyses. Importantly, buffer-area LULC change magnitude was frequently much greater than footprint-level change. Surface-water extent in buffers increased by 14–35x the footprint rate and forest decreased by 2–9x. Development in buffering areas increased by 2–4x the footprint-rate in Chicago and Peoria area footprints but was similar to the change rate in the St. Louis area footprint. The LULC buffer-area change varied in waterbody size, with the greatest change typically occurring in the smallest waters (e.g., <0.1 ha). These novel analyses suggest that surface-water buffer LULC change is occurring more rapidly than footprint-level change, likely modifying the hydrology, water quality, and biotic integrity of existing water resources, as well as potentially affecting down-gradient, watershed-scale storages and flows of water, solutes, and particulate matter. Full article

(This article belongs to the Special Issue Advancements in Remote Sensing of Land Surface Change)

► Show Figures

Graphical abstract

40 pages, 3499 KB

Open AccessFeature PaperEditor’s ChoiceReview

The Spatial and Spectral Resolution of ASTER Infrared Image Data: A Paradigm Shift in Volcanological Remote Sensing

by Michael S. Ramsey and Ian T.W. Flynn

Remote Sens. 2020, 12(4), 738; https://doi.org/10.3390/rs12040738 - 23 Feb 2020

Cited by 32 | Viewed by 10979

Abstract

During the past two decades, the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) instrument on the Terra satellite has acquired nearly 320,000 scenes of the world’s volcanoes. This is ~10% of the data in the global ASTER archive. Many of these scenes [...] Read more.

During the past two decades, the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) instrument on the Terra satellite has acquired nearly 320,000 scenes of the world’s volcanoes. This is ~10% of the data in the global ASTER archive. Many of these scenes captured volcanic activity at never before seen spatial and spectral scales, particularly in the thermal infrared (TIR) region. Despite this large archive of data, the temporal resolution of ASTER is simply not adequate to understand ongoing eruptions and assess the hazards to local populations in near real time. However, programs designed to integrate ASTER into a volcanic data sensor web have greatly improved the cadence of the data (in some cases, to as many as 3 scenes in 48 h). This frequency can inform our understanding of what is possible with future systems collecting similar data on the daily or hourly time scales. Here, we present the history of ASTER’s contributions to volcanology, highlighting unique aspects of the instrument and its data. The ASTER archive was mined to provide statistics including the number of observations with volcanic activity, its type, and the average cloud cover. These were noted for more than 2000 scenes over periods of 1, 5 and 20 years. Full article

(This article belongs to the Special Issue ASTER 20th Anniversary)

► Show Figures

Graphical abstract

20 pages, 4249 KB

Open AccessEditor’s ChoiceArticle

A High-Resolution Global Map of Giant Kelp (Macrocystis pyrifera) Forests and Intertidal Green Algae (Ulvophyceae) with Sentinel-2 Imagery

by Alejandra Mora-Soto, Mauricio Palacios, Erasmo C. Macaya, Iván Gómez, Pirjo Huovinen, Alejandro Pérez-Matus, Mary Young, Neil Golding, Martin Toro, Mohammad Yaqub and Marc Macias-Fauria

Remote Sens. 2020, 12(4), 694; https://doi.org/10.3390/rs12040694 - 20 Feb 2020

Cited by 89 | Viewed by 20825

Abstract

Giant kelp (Macrocystis pyrifera) is the most widely distributed kelp species on the planet, constituting one of the richest and most productive ecosystems on Earth, but detailed information on its distribution is entirely missing in some marine ecoregions, especially in the [...] Read more.

Giant kelp (Macrocystis pyrifera) is the most widely distributed kelp species on the planet, constituting one of the richest and most productive ecosystems on Earth, but detailed information on its distribution is entirely missing in some marine ecoregions, especially in the high latitudes of the Southern Hemisphere. Here, we present an algorithm based on a series of filter thresholds to detect giant kelp employing Sentinel-2 imagery. Given the overlap between the reflectances of giant kelp and intertidal green algae (Ulvophyceae), the latter are also detected on shallow rocky intertidal areas. The kelp filter algorithm was applied separately to vegetation indices, the Floating Algae Index (FAI), the Normalised Difference Vegetation Index (NDVI), and a novel formula (the Kelp Difference, KD). Training data from previously surveyed kelp forests and other coastal and ocean features were used to identify reflectance threshold values. This procedure was validated with independent field data collected with UAV imagery at a high spatial resolution and point-georeferenced sites at a low spatial resolution. When comparing UAV with Sentinel data (high-resolution validation), an average overall accuracy ≥ 0.88 and Cohen’s kappa ≥ 0.64 coefficients were found in all three indices for canopies reaching the surface with extensions greater than 1 hectare, with the KD showing the highest average kappa score (0.66). Measurements between previously surveyed georeferenced points and remotely-sensed kelp grid cells (low-resolution validation) showed that 66% of the georeferenced points had grid cells indicating kelp presence within a linear distance of 300 m. We employed the KD in our kelp filter algorithm to estimate the global extent of giant kelp and intertidal green algae per marine ecoregion and province, producing a high-resolution global map of giant kelp and intertidal green algae, powered by Google Earth Engine. Full article

(This article belongs to the Special Issue Drone-Based Ecological Conservation)

► Show Figures

Graphical abstract

13 pages, 3191 KB

Open AccessEditor’s ChoiceLetter

Error Estimation of Pathfinder Version 5.3 Level-3C SST Using Extended Triple Collocation Analysis

by Korak Saha, Prasanjit Dash, Xuepeng Zhao and Huai-min Zhang

Remote Sens. 2020, 12(4), 590; https://doi.org/10.3390/rs12040590 - 11 Feb 2020

Cited by 17 | Viewed by 5173

Abstract

Sea Surface Temperature (SST) is an essential climate variable (ECV) for monitoring the state and detecting changes in the climate. The concept of ECVs, developed by the Global Climate Observing System (GCOS) program of the World Meteorological Organization (WMO), has been broadly adopted [...] Read more.

Sea Surface Temperature (SST) is an essential climate variable (ECV) for monitoring the state and detecting changes in the climate. The concept of ECVs, developed by the Global Climate Observing System (GCOS) program of the World Meteorological Organization (WMO), has been broadly adopted in worldwide science and policy circles Besides being a climate change indicator, the global SST field is an essential input for atmospheric models, air-sea exchange studies, understanding marine ecosystems, operational weather, and ocean forecasting, military and defense operations, tourism, and fisheries research. It is, therefore, critical to understand the errors associated with SST measurements from both in situ measurements and satellite observations. The customary way of validating a satellite SST is to compare it with in situ measured SSTs. This method, however, will have inaccuracies due to uncertainties involving both types of measurements. A triple collocation (TC) error analysis can be implemented on three mutually independent error-prone measurements to estimate the root-mean-square error (RMSE) of each measurement. In this study, the error characterization for the Pathfinder SST version 5.3 (PF53) dataset is performed using an extended TC (ETC) method and reported to be in the range of 0.31 to 0.37 K. These values are reasonable, as is evident from corresponding very high (~0.98) unbiased signal-to-noise ratio (SNR) values. Full article

(This article belongs to the Special Issue Advances in Retrieval, Operationalization, Monitoring and Application of Sea Surface Temperature)

► Show Figures

Graphical abstract

22 pages, 9201 KB

Open AccessEditor’s ChoiceArticle

Mapping the Land Cover of Africa at 10 m Resolution from Multi-Source Remote Sensing Data with Google Earth Engine

by Qingyu Li, Chunping Qiu, Lei Ma, Michael Schmitt and Xiao Xiang Zhu

Remote Sens. 2020, 12(4), 602; https://doi.org/10.3390/rs12040602 - 11 Feb 2020

Cited by 81 | Viewed by 14372

Abstract

The remote sensing based mapping of land cover at extensive scales, e.g., of whole continents, is still a challenging task because of the need for sophisticated pipelines that combine every step from data acquisition to land cover classification. Utilizing the Google Earth Engine [...] Read more.

The remote sensing based mapping of land cover at extensive scales, e.g., of whole continents, is still a challenging task because of the need for sophisticated pipelines that combine every step from data acquisition to land cover classification. Utilizing the Google Earth Engine (GEE), which provides a catalog of multi-source data and a cloud-based environment, this research generates a land cover map of the whole African continent at 10 m resolution. This land cover map could provide a large-scale base layer for a more detailed local climate zone mapping of urban areas, which lie in the focus of interest of many studies. In this regard, we provide a free download link for our land cover maps of African cities at the end of this paper. It is shown that our product has achieved an overall accuracy of 81% for five classes, which is superior to the existing 10 m land cover product FROM-GLC10 in detecting urban class in city areas and identifying the boundaries between trees and low plants in rural areas. The best data input configurations are carefully selected based on a comparison of results from different input sources, which include Sentinel-2, Landsat-8, Global Human Settlement Layer (GHSL), Night Time Light (NTL) Data, Shuttle Radar Topography Mission (SRTM), and MODIS Land Surface Temperature (LST). We provide a further investigation of the importance of individual features derived from a Random Forest (RF) classifier. In order to study the influence of sampling strategies on the land cover mapping performance, we have designed a transferability analysis experiment, which has not been adequately addressed in the current literature. In this experiment, we test whether trained models from several cities contain valuable information to classify a different city. It was found that samples of the urban class have better reusability than those of other natural land cover classes, i.e., trees, low plants, bare soil or sand, and water. After experimental evaluation of different land cover classes across different cities, we conclude that continental land cover mapping results can be considerably improved when training samples of natural land cover classes are collected and combined from areas covering each Köppen climate zone. Full article

(This article belongs to the Special Issue Regional and Global Land Cover Mapping)

► Show Figures

Figure 1

23 pages, 5267 KB

Open AccessEditor’s ChoiceArticle

Towards Routine Mapping of Shallow Bathymetry in Environments with Variable Turbidity: Contribution of Sentinel-2A/B Satellites Mission

by Isabel Caballero and Richard P. Stumpf

Remote Sens. 2020, 12(3), 451; https://doi.org/10.3390/rs12030451 - 1 Feb 2020

Cited by 109 | Viewed by 9778

Abstract

Satellite-Derived Bathymetry (SDB) has significant potential to enhance our knowledge of Earth’s coastal regions. However, SDB still has limitations when applied to the turbid, but optically shallow, nearshore regions that encompass large areas of the world’s coastal zone. Turbid water produces false shoaling [...] Read more.

Satellite-Derived Bathymetry (SDB) has significant potential to enhance our knowledge of Earth’s coastal regions. However, SDB still has limitations when applied to the turbid, but optically shallow, nearshore regions that encompass large areas of the world’s coastal zone. Turbid water produces false shoaling in the imagery, constraining SDB for its routine application. This paper provides a framework that enables us to derive valid SDB over moderately turbid environments by using the high revisit time (5-day) of the Sentinel-2A/B twin mission from the Copernicus programme. The proposed methodology incorporates a robust atmospheric correction, a multi-scene compositing method to reduce the impact of turbidity, and a switching model to improve mapping in shallow water. Two study sites in United States are explored due to their varying water transparency conditions. Our results show that the approach yields accurate SDB, with median errors of under 0.5 m for depths 0–13 m when validated with lidar surveys, errors that favorably compare to uses of SDB in clear water. The approach allows for the semi-automated creation of bathymetric maps at 10 m spatial resolution, with manual intervention potentially limited only to the calibration to the absolute SDB. It also returns turbidity data to indicate areas that may still have residual shoaling bias. Because minimal in-situ information is required, this computationally-efficient technique has the potential for automated implementation, allowing rapid and repeated application in more environments than most existing methods, thereby helping with a range of issues in coastal research, management, and navigation. Full article

(This article belongs to the Special Issue Coastal Waters Monitoring Using Remote Sensing Technology)

► Show Figures

Graphical abstract

29 pages, 14114 KB

Open AccessEditor’s ChoiceArticle

How Well Do Deep Learning-Based Methods for Land Cover Classification and Object Detection Perform on High Resolution Remote Sensing Imagery?

by Xin Zhang, Liangxiu Han, Lianghao Han and Liang Zhu

Remote Sens. 2020, 12(3), 417; https://doi.org/10.3390/rs12030417 - 28 Jan 2020

Cited by 145 | Viewed by 16225

Abstract

Land cover information plays an important role in mapping ecological and environmental changes in Earth’s diverse landscapes for ecosystem monitoring. Remote sensing data have been widely used for the study of land cover, enabling efficient mapping of changes of the Earth surface from [...] Read more.

Land cover information plays an important role in mapping ecological and environmental changes in Earth’s diverse landscapes for ecosystem monitoring. Remote sensing data have been widely used for the study of land cover, enabling efficient mapping of changes of the Earth surface from Space. Although the availability of high-resolution remote sensing imagery increases significantly every year, traditional land cover analysis approaches based on pixel and object levels are not optimal. Recent advancement in deep learning has achieved remarkable success on image recognition field and has shown potential in high spatial resolution remote sensing applications, including classification and object detection. In this paper, a comprehensive review on land cover classification and object detection approaches using high resolution imagery is provided. Through two case studies, we demonstrated the applications of the state-of-the-art deep learning models to high spatial resolution remote sensing data for land cover classification and object detection and evaluated their performances against traditional approaches. For a land cover classification task, the deep-learning-based methods provide an end-to-end solution by using both spatial and spectral information. They have shown better performance than the traditional pixel-based method, especially for the categories of different vegetation. For an objective detection task, the deep-learning-based object detection method achieved more than 98% accuracy in a large area; its high accuracy and efficiency could relieve the burden of the traditional, labour-intensive method. However, considering the diversity of remote sensing data, more training datasets are required in order to improve the generalisation and the robustness of deep learning-based models. Full article

(This article belongs to the Special Issue Advanced Machine Learning Techniques for High-Resolution Remote Sensing Data Analysis)

► Show Figures

Graphical abstract

29 pages, 6546 KB

Open AccessEditor’s ChoiceArticle

LiCSBAS: An Open-Source InSAR Time Series Analysis Package Integrated with the LiCSAR Automated Sentinel-1 InSAR Processor

by Yu Morishita, Milan Lazecky, Tim J. Wright, Jonathan R. Weiss, John R. Elliott and Andy Hooper

Remote Sens. 2020, 12(3), 424; https://doi.org/10.3390/rs12030424 - 28 Jan 2020

Cited by 288 | Viewed by 39124

Abstract

For the past five years, the 2-satellite Sentinel-1 constellation has provided abundant and useful Synthetic Aperture Radar (SAR) data, which have the potential to reveal global ground surface deformation at high spatial and temporal resolutions. However, for most users, fully exploiting the large [...] Read more.

For the past five years, the 2-satellite Sentinel-1 constellation has provided abundant and useful Synthetic Aperture Radar (SAR) data, which have the potential to reveal global ground surface deformation at high spatial and temporal resolutions. However, for most users, fully exploiting the large amount of associated data is challenging, especially over wide areas. To help address this challenge, we have developed LiCSBAS, an open-source SAR interferometry (InSAR) time series analysis package that integrates with the automated Sentinel-1 InSAR processor (LiCSAR). LiCSBAS utilizes freely available LiCSAR products, and users can save processing time and disk space while obtaining the results of InSAR time series analysis. In the LiCSBAS processing scheme, interferograms with many unwrapping errors are automatically identified by loop closure and removed. Reliable time series and velocities are derived with the aid of masking using several noise indices. The easy implementation of atmospheric corrections to reduce noise is achieved with the Generic Atmospheric Correction Online Service for InSAR (GACOS). Using case studies in southern Tohoku and the Echigo Plain, Japan, we demonstrate that LiCSBAS applied to LiCSAR products can detect both large-scale (>100 km) and localized (~km) relative displacements with an accuracy of <1 cm/epoch and ~2 mm/yr. We detect displacements with different temporal characteristics, including linear, periodic, and episodic, in Niigata, Ojiya, and Sanjo City, respectively. LiCSBAS and LiCSAR products facilitate greater exploitation of globally available and abundant SAR datasets and enhance their applications for scientific research and societal benefit. Full article

(This article belongs to the Special Issue Scaling-Up Deformation Monitoring and Analysis)

► Show Figures

Graphical abstract

22 pages, 5641 KB

Open AccessFeature PaperEditor’s ChoiceArticle

Comparison of Machine Learning Methods Applied to SAR Images for Forest Classification in Mediterranean Areas

by Alessandro Lapini, Simone Pettinato, Emanuele Santi, Simonetta Paloscia, Giacomo Fontanelli and Andrea Garzelli

Remote Sens. 2020, 12(3), 369; https://doi.org/10.3390/rs12030369 - 22 Jan 2020

Cited by 47 | Viewed by 6062

Abstract

In this paper, multifrequency synthetic aperture radar (SAR) images from ALOS/PALSAR, ENVISAT/ASAR and Cosmo-SkyMed sensors were studied for forest classification in a test area in Central Italy (San Rossore), where detailed in-situ measurements were available. A preliminary discrimination of the main land cover [...] Read more.

In this paper, multifrequency synthetic aperture radar (SAR) images from ALOS/PALSAR, ENVISAT/ASAR and Cosmo-SkyMed sensors were studied for forest classification in a test area in Central Italy (San Rossore), where detailed in-situ measurements were available. A preliminary discrimination of the main land cover classes and forest types was carried out by exploiting the synergy among L-, C- and X-bands and different polarizations. SAR data were preliminarily inspected to assess the capabilities of discriminating forest from non-forest and separating broadleaf from coniferous forests. The temporal average backscattering coefficient (

\bar{σ}

°) was computed for each sensor-polarization pair and labeled on a pixel basis according to the reference map. Several classification methods based on the machine learning framework were applied and validated considering different features, in order to highlight the contribution of bands and polarizations, as well as to assess the classifiers’ performance. The experimental results indicate that the different surface types are best identified by using all bands, followed by joint L- and X-bands. In the former case, the best overall average accuracy (83.1%) is achieved by random forest classification. Finally, the classification maps on class edges are discussed to highlight the misclassification errors. Full article

(This article belongs to the Special Issue Mediterranean Forest Monitoring Using Optical and Microwave Remote Sensing)

► Show Figures

Graphical abstract

23 pages, 27978 KB

Open AccessEditor’s ChoiceArticle

Mapping Landslides on EO Data: Performance of Deep Learning Models vs. Traditional Machine Learning Models

by Nikhil Prakash, Andrea Manconi and Simon Loew

Remote Sens. 2020, 12(3), 346; https://doi.org/10.3390/rs12030346 - 21 Jan 2020

Cited by 229 | Viewed by 14176

Abstract

Mapping landslides using automated methods is a challenging task, which is still largely done using human efforts. Today, the availability of high-resolution EO data products is increasing exponentially, and one of the targets is to exploit this data source for the rapid generation [...] Read more.

Mapping landslides using automated methods is a challenging task, which is still largely done using human efforts. Today, the availability of high-resolution EO data products is increasing exponentially, and one of the targets is to exploit this data source for the rapid generation of landslide inventory. Conventional methods like pixel-based and object-based machine learning strategies have been studied extensively in the last decade. In addition, recent advances in CNN (convolutional neural network), a type of deep-learning method, has been widely successful in extracting information from images and have outperformed other conventional learning methods. In the last few years, there have been only a few attempts to adapt CNN for landslide mapping. In this study, we introduce a modified U-Net model for semantic segmentation of landslides at a regional scale from EO data using ResNet34 blocks for feature extraction. We also compare this with conventional pixel-based and object-based methods. The experiment was done in Douglas County, a study area selected in the south of Portland in Oregon, USA, and landslide inventory extracted from SLIDO (Statewide Landslide Information Database of Oregon) was considered as the ground truth. Landslide mapping is an imbalanced learning problem with very limited availability of training data. Our network was trained on a combination of focal Tversky loss and cross-entropy loss functions using augmented image tiles sampled from a selected training area. The deep-learning method was observed to have a better performance than the conventional methods with an MCC (Matthews correlation coefficient) score of

0.495

and a POD (probability of detection) rate of

0.72

. Full article

(This article belongs to the Special Issue Landslide Monitoring, Susceptibility, Hazard Assessment and Prediction with Remotely Sensed Big Data)

► Show Figures

Graphical abstract

22 pages, 14105 KB

Open AccessEditor’s ChoiceArticle

Integrating Remote Sensing and Street View Images to Quantify Urban Forest Ecosystem Services

by Elena Barbierato, Iacopo Bernetti, Irene Capecchi and Claudio Saragosa

Remote Sens. 2020, 12(2), 329; https://doi.org/10.3390/rs12020329 - 19 Jan 2020

Cited by 54 | Viewed by 8615

Abstract

There is an urgent need for holistic tools to assess the health impacts of climate change mitigation and adaptation policies relating to increasing public green spaces. Urban vegetation provides numerous ecosystem services on a local scale and is therefore a potential adaptation strategy [...] Read more.

There is an urgent need for holistic tools to assess the health impacts of climate change mitigation and adaptation policies relating to increasing public green spaces. Urban vegetation provides numerous ecosystem services on a local scale and is therefore a potential adaptation strategy that can be used in an era of global warming to offset the increasing impacts of human activity on urban environments. In this study, we propose a set of urban green ecological metrics that can be used to evaluate urban green ecosystem services. The metrics were derived from two complementary surveys: a traditional remote sensing survey of multispectral images and Laser Imaging Detection and Ranging (LiDAR) data, and a survey using proximate sensing through images made available by the Google Street View database. In accordance with previous studies, two classes of metrics were calculated: greenery at lower and higher elevations than building facades. In the last phase of the work, the metrics were applied to city blocks, and a spatially constrained clustering methodology was employed. Homogeneous areas were identified in relation to the urban greenery characteristics. The proposed methodology represents the development of a geographic information system that can be used by public administrators and urban green designers to create and maintain urban public forests. Full article

(This article belongs to the Special Issue Remote Sensing in Applications of Geoinformation)

► Show Figures

Graphical abstract

22 pages, 33791 KB

Open AccessEditor’s ChoiceArticle

Predicting Forest Cover in Distinct Ecosystems: The Potential of Multi-Source Sentinel-1 and -2 Data Fusion

by Kai Heckel, Marcel Urban, Patrick Schratz, Miguel D. Mahecha and Christiane Schmullius

Remote Sens. 2020, 12(2), 302; https://doi.org/10.3390/rs12020302 - 17 Jan 2020

Cited by 61 | Viewed by 10325

Abstract

The fusion of microwave and optical data sets is expected to provide great potential for the derivation of forest cover around the globe. As Sentinel-1 and Sentinel-2 are now both operating in twin mode, they can provide an unprecedented data source to build [...] Read more.

The fusion of microwave and optical data sets is expected to provide great potential for the derivation of forest cover around the globe. As Sentinel-1 and Sentinel-2 are now both operating in twin mode, they can provide an unprecedented data source to build dense spatial and temporal high-resolution time series across a variety of wavelengths. This study investigates (i) the ability of the individual sensors and (ii) their joint potential to delineate forest cover for study sites in two highly varied landscapes located in Germany (temperate dense mixed forests) and South Africa (open savanna woody vegetation and forest plantations). We used multi-temporal Sentinel-1 and single time steps of Sentinel-2 data in combination to derive accurate forest/non-forest (FNF) information via machine-learning classifiers. The forest classification accuracies were 90.9% and 93.2% for South Africa and Thuringia, respectively, estimated while using autocorrelation corrected spatial cross-validation (CV) for the fused data set. Sentinel-1 only classifications provided the lowest overall accuracy of 87.5%, while Sentinel-2 based classifications led to higher accuracies of 91.9%. Sentinel-2 short-wave infrared (SWIR) channels, biophysical parameters (Leaf Area Index (LAI), and Fraction of Absorbed Photosynthetically Active Radiation (FAPAR)) and the lower spectrum of the Sentinel-1 synthetic aperture radar (SAR) time series were found to be most distinctive in the detection of forest cover. In contrast to homogenous forests sites, Sentinel-1 time series information improved forest cover predictions in open savanna-like environments with heterogeneous regional features. The presented approach proved to be robust and it displayed the benefit of fusing optical and SAR data at high spatial resolution. Full article

(This article belongs to the Special Issue Advances in Remote Sensing for Global Forest Monitoring)

► Show Figures

Graphical abstract

18 pages, 5408 KB

Open AccessEditor’s ChoiceArticle

Harmonization of Landsat and Sentinel 2 for Crop Monitoring in Drought Prone Areas: Case Studies of Ninh Thuan (Vietnam) and Bekaa (Lebanon)

by Minh D. Nguyen, Oscar M. Baez-Villanueva, Duong D. Bui, Phong T. Nguyen and Lars Ribbe

Remote Sens. 2020, 12(2), 281; https://doi.org/10.3390/rs12020281 - 15 Jan 2020

Cited by 83 | Viewed by 14524

Abstract

Proper satellite-based crop monitoring applications at the farm-level often require near-daily imagery at medium to high spatial resolution. The combination of data from different ongoing satellite missions Sentinel 2 (ESA) and Landsat 7/8 (NASA) provides this unprecedented opportunity at a global scale; however, [...] Read more.

Proper satellite-based crop monitoring applications at the farm-level often require near-daily imagery at medium to high spatial resolution. The combination of data from different ongoing satellite missions Sentinel 2 (ESA) and Landsat 7/8 (NASA) provides this unprecedented opportunity at a global scale; however, this is rarely implemented because these procedures are data demanding and computationally intensive. This study developed a robust stream processing for the harmonization of Landsat 7, Landsat 8 and Sentinel 2 in the Google Earth Engine cloud platform, connecting the benefit of coherent data structure, built-in functions and computational power in the Google Cloud. The harmonized surface reflectance images were generated for two agricultural schemes in Bekaa (Lebanon) and Ninh Thuan (Vietnam) during 2018–2019. We evaluated the performance of several pre-processing steps needed for the harmonization including the image co-registration, Bidirectional Reflectance Distribution Functions correction, topographic correction, and band adjustment. We found that the misregistration between Landsat 8 and Sentinel 2 images varied from 10 m in Ninh Thuan (Vietnam) to 32 m in Bekaa (Lebanon), and posed a great impact on the quality of the final harmonized data set if not treated. Analysis of a pair of overlapped L8-S2 images over the Bekaa region showed that, after the harmonization, all band-to-band spatial correlations were greatly improved. Finally, we demonstrated an application of the dense harmonized data set for crop mapping and monitoring. An harmonic (Fourier) analysis was applied to fit the detected unimodal, bimodal and trimodal shapes in the temporal NDVI patterns during one crop year in Ninh Thuan province. The derived phase and amplitude values of the crop cycles were combined with max-NDVI as an R-G-B false composite image. The final image was able to highlight croplands in bright colors (high phase and amplitude), while the non-crop areas were shown with grey/dark (low phase and amplitude). The harmonized data sets (with 30 m spatial resolution) along with the Google Earth Engine scripts used are provided for public use. Full article

► Show Figures

Graphical abstract

26 pages, 6611 KB

Open AccessEditor’s ChoiceArticle

Evaluation of Coherent and Incoherent Landslide Detection Methods Based on Synthetic Aperture Radar for Rapid Response: A Case Study for the 2018 Hokkaido Landslides

by Jungkyo Jung and Sang-Ho Yun

Remote Sens. 2020, 12(2), 265; https://doi.org/10.3390/rs12020265 - 13 Jan 2020

Cited by 49 | Viewed by 5812

Abstract

Damage mapping using Synthetic Aperture Radar (SAR) imagery has been studied in recent decades to support rapid response to natural disasters. Many researches have been developing coherent and incoherent change detection. However, their performances can vary depending on the types of the damages, [...] Read more.

Damage mapping using Synthetic Aperture Radar (SAR) imagery has been studied in recent decades to support rapid response to natural disasters. Many researches have been developing coherent and incoherent change detection. However, their performances can vary depending on the types of the damages, the characteristics of the scatterers and the corresponding capability of algorithms. In particular, the coherence-based methods have been used as promising detectors over urban areas where high coherences are observed, but their detection accuracies still remain controversial over the area where low coherences are mainly observed such as the 2018 Hokkaido landslides. In order to understand the characteristics of landslide (damage) detectors for low-coherence areas and find an alternative and complementary method, we designed the coherence difference, coherence normalized difference, log-ratio, intensity correlation difference, and normalized differences of the intensity correlation assuming limited availability of dataset, and also developed multi-temporal algorithms using the coherence, intensity, and intensity correlation. They were tested and evaluated using multiple polygons extracted from aerial photos. We were able to observe that the multi-temporal intensity correlation method has the potential to detect the landslides over the low coherence region and all types of land uses. Full article

(This article belongs to the Section Environmental Remote Sensing)

► Show Figures

Graphical abstract

17 pages, 7669 KB

Open AccessEditor’s ChoiceArticle

Cloud Removal with Fusion of High Resolution Optical and SAR Images Using Generative Adversarial Networks

by Jianhao Gao, Qiangqiang Yuan, Jie Li, Hai Zhang and Xin Su

Remote Sens. 2020, 12(1), 191; https://doi.org/10.3390/rs12010191 - 5 Jan 2020

Cited by 142 | Viewed by 11866

Abstract

The existence of clouds is one of the main factors that contributes to missing information in optical remote sensing images, restricting their further applications for Earth observation, so how to reconstruct the missing information caused by clouds is of great concern. Inspired by [...] Read more.

The existence of clouds is one of the main factors that contributes to missing information in optical remote sensing images, restricting their further applications for Earth observation, so how to reconstruct the missing information caused by clouds is of great concern. Inspired by the image-to-image translation work based on convolutional neural network model and the heterogeneous information fusion thought, we propose a novel cloud removal method in this paper. The approach can be roughly divided into two steps: in the first step, a specially designed convolutional neural network (CNN) translates the synthetic aperture radar (SAR) images into simulated optical images in an object-to-object manner; in the second step, the simulated optical image, together with the SAR image and the optical image corrupted by clouds, is fused to reconstruct the corrupted area by a generative adversarial network (GAN) with a particular loss function. Between the first step and the second step, the contrast and luminance of the simulated optical image are randomly altered to make the model more robust. Two simulation experiments and one real-data experiment are conducted to confirm the effectiveness of the proposed method on Sentinel 1/2, GF 2/3 and airborne SAR/optical data. The results demonstrate that the proposed method outperforms state-of-the-art algorithms that also employ SAR images as auxiliary data. Full article

(This article belongs to the Section Remote Sensing Image Processing)

► Show Figures

Graphical abstract

19 pages, 13416 KB

Open AccessEditor’s ChoiceArticle

Antarctic Supraglacial Lake Detection Using Landsat 8 and Sentinel-2 Imagery: Towards Continental Generation of Lake Volumes

by Mahsa Moussavi, Allen Pope, Anna Ruth W. Halberstadt, Luke D. Trusel, Leanne Cioffi and Waleed Abdalati

Remote Sens. 2020, 12(1), 134; https://doi.org/10.3390/rs12010134 - 1 Jan 2020

Cited by 81 | Viewed by 12348

Abstract

Melt and supraglacial lakes are precursors to ice shelf collapse and subsequent accelerated ice sheet mass loss. We used data from the Landsat 8 and Sentinel-2 satellites to develop a threshold-based method for detection of lakes found on the Antarctic ice shelves, calculate [...] Read more.

Melt and supraglacial lakes are precursors to ice shelf collapse and subsequent accelerated ice sheet mass loss. We used data from the Landsat 8 and Sentinel-2 satellites to develop a threshold-based method for detection of lakes found on the Antarctic ice shelves, calculate their depths and thus their volumes. To achieve this, we focus on four key areas: the Amery, Roi Baudouin, Nivlisen, and Riiser-Larsen ice shelves, which are all characterized by extensive surface meltwater features. To validate our products, we compare our results against those obtained by an independent method based on a supervised classification scheme (e.g., Random Forest algorithm). Additional verification is provided by manual inspection of results for nearly 1000 Landsat 8 and Sentinel-2 images. Our dual-sensor approach will enable constructing high-resolution time series of lake volumes. Therefore, to ensure interoperability between the two datasets, we evaluate depths from contemporaneous Landsat 8 and Sentinel-2 image pairs. Our assessments point to a high degree of correspondence, producing an average R² value of 0.85, no bias, and an average RMSE of 0.2 m. We demonstrate our method’s ability to characterize lake evolution by presenting first evidence of drainage events outside of the Antarctic Peninsula on the Amery Ice shelf. The methods presented here pave the way to upscaling throughout the Landsat 8 and Sentinel-2 observational record across Antarctica to produce a first-ever continental dataset of supraglacial lake volumes. Such a dataset will improve our understanding of the influence of surface hydrology on ice shelf stability, and thus, future projections of Antarctica’s contribution to sea level rise. Full article

(This article belongs to the Section Environmental Remote Sensing)

► Show Figures

Graphical abstract

15 pages, 2238 KB

Open AccessEditor’s ChoiceArticle

Potential of Night-Time Lights to Measure Regional Inequality

by Kinga Ivan, Iulian-Horia Holobâcă, József Benedek and Ibolya Török

Remote Sens. 2020, 12(1), 33; https://doi.org/10.3390/rs12010033 - 20 Dec 2019

Cited by 56 | Viewed by 9171

Abstract

Night-time lights satellite images provide a new opportunity to measure regional inequality in real-time by developing the Night Light Development Index (NLDI). The NLDI was extracted using the Gini coefficient approach based on population and night light spatial distribution in Romania. Night-time light [...] Read more.

Night-time lights satellite images provide a new opportunity to measure regional inequality in real-time by developing the Night Light Development Index (NLDI). The NLDI was extracted using the Gini coefficient approach based on population and night light spatial distribution in Romania. Night-time light data were calculated using a grid with a 0.15 km² area, based on Defense Meteorological Satellite Program (DMSP) /Operational Linescan System (OLS satellite imagery for the 1992–2013 period and based on the National Polar-orbiting Partnership–Visible Infrared Imaging Radiometer Suite (NPP-VIIRS) satellite imagery for the 2014–2018 period. Two population density grids were created at the level of equal cells (0.15 km²) using ArcGIS and PostgreSQL software, and census data from 1992 and 2011. Subsequently, based on this data and using the Gini index approach, the Night Light Development Index (NLDI) was calculated within the MATLAB software. The NLDI was obtained for 42 administrative counties (nomenclature of territorial units for statistics level 3 (NUTS-3 units)) for the 1992–2018 period. The statistical relationship between the NLDI and the socio-economic, demographic, and geographic variables highlighted a strong indirect relationship with local tax income and gross domestic product (GDP) per capita. The polynomial model proved to be better in estimating income based on the NLDI and R² coefficients showed a significant improvement in total variation explained compared to the linear regression model. The NLDI calculated on the basis of night-time lights satellite images proved to be a good proxy for measuring regional inequalities. Therefore, it can play a crucial role in monitoring the progress made in the implementation of Sustainable Development Goal 10 (reduced inequalities). Full article

(This article belongs to the Special Issue EO Solutions to Support Countries Implementing the SDGs)

► Show Figures

Figure 1

18 pages, 6109 KB

Open AccessEditor’s ChoiceArticle

Feasibility of Burned Area Mapping Based on ICESAT−2 Photon Counting Data

by Meng Liu, Sorin Popescu and Lonesome Malambo

Remote Sens. 2020, 12(1), 24; https://doi.org/10.3390/rs12010024 - 19 Dec 2019

Cited by 37 | Viewed by 6313

Abstract

Accurately mapping burned areas is crucial for the analysis of carbon emissions and wildfire risk as well as understanding the effects of climate change on forest structure. Burned areas have predominantly been mapped using optical remote sensing images. However, the structural changes due [...] Read more.

Accurately mapping burned areas is crucial for the analysis of carbon emissions and wildfire risk as well as understanding the effects of climate change on forest structure. Burned areas have predominantly been mapped using optical remote sensing images. However, the structural changes due to fire also offer opportunities for mapping burned areas using three-dimensional (3D) datasets such as Light detection and ranging (LiDAR). This study focuses on the feasibility of using photon counting LiDAR data from National Aeronautics and Space Administration’s (NASA) Ice, Cloud, and land Elevation Satellite-2 (ICESat−2) mission to differentiate vegetation structure in burned and unburned areas and ultimately classify burned areas along mapped ground tracks. The ICESat−2 mission (launched in September 2018) provides datasets such as geolocated photon data (ATL03), which comprises precise latitude, longitude and elevation of each point where a photon interacts with land surface, and derivative products such as the Land Water Vegetation Elevation product (ATL08), which comprises estimated terrain and canopy height information. For analysis, 24 metrics such as the average, median and standard deviation of canopy height were derived from ATL08 data over forests burned by recent fires in 2018 in northern California and western New Mexico. A reference burn map was derived from Sentinel−2 images based on the differenced Normalized Burn Ratio (dNBR) index. A landcover map based on Sentinel−2 images was employed to remove non-forest classes. Landsat 8 based dNBR image and landcover map were also used for comparison. Next, ICESat−2 data of forest samples were classified into burned and unburned ATL08 100-m segments by both Random Forest classification and logistic regression. Both Sentinel−2 derived and Landsat 8 derived ATL08 samples got high classification accuracy, 83% versus 76%. Moreover, the resulting classification accuracy by Random Forest and logistic regression reached 83% and 74%, respectively. Among the 24 ICESat−2 metrics, apparent surface reflectance and the number of canopy photons were the most important. Furthermore, burn severity of each ATL08 segment was also estimated with Random Forest regression. R² of predicted burn severity to observed dNBR is 0.61 with significant linear relationship and moderate correlation (r = 0.78). Overall, the reasonably high accuracies achieved in this study demonstrate the feasibility of employing ICESat−2 data in burned forest classification, opening avenues for improved estimation of burned biomass and carbon emissions from a 3D perspective. Full article

(This article belongs to the Special Issue Lidar Remote Sensing of Forest Structure, Biomass and Dynamics)

► Show Figures

Graphical abstract

19 pages, 1016 KB

Open AccessFeature PaperEditor’s ChoiceArticle

Remote Sensing Big Data Classification with High Performance Distributed Deep Learning

by Rocco Sedona, Gabriele Cavallaro, Jenia Jitsev, Alexandre Strube, Morris Riedel and Jón Atli Benediktsson

Remote Sens. 2019, 11(24), 3056; https://doi.org/10.3390/rs11243056 - 17 Dec 2019

Cited by 38 | Viewed by 8217

Abstract

High-Performance Computing (HPC) has recently been attracting more attention in remote sensing applications due to the challenges posed by the increased amount of open data that are produced daily by Earth Observation (EO) programs. The unique parallel computing environments and programming techniques that [...] Read more.

High-Performance Computing (HPC) has recently been attracting more attention in remote sensing applications due to the challenges posed by the increased amount of open data that are produced daily by Earth Observation (EO) programs. The unique parallel computing environments and programming techniques that are integrated in HPC systems are able to solve large-scale problems such as the training of classification algorithms with large amounts of Remote Sensing (RS) data. This paper shows that the training of state-of-the-art deep Convolutional Neural Networks (CNNs) can be efficiently performed in distributed fashion using parallel implementation techniques on HPC machines containing a large number of Graphics Processing Units (GPUs). The experimental results confirm that distributed training can drastically reduce the amount of time needed to perform full training, resulting in near linear scaling without loss of test accuracy. Full article

(This article belongs to the Special Issue Analysis of Big Data in Remote Sensing)

► Show Figures

Graphical abstract

18 pages, 10412 KB

Open AccessEditor’s ChoiceArticle

TanDEM-X Forest Mapping Using Convolutional Neural Networks

by Antonio Mazza, Francescopaolo Sica, Paola Rizzoli and Giuseppe Scarpa

Remote Sens. 2019, 11(24), 2980; https://doi.org/10.3390/rs11242980 - 12 Dec 2019

Cited by 37 | Viewed by 4798

Abstract

In this work, we face the problem of forest mapping from TanDEM-X data by means of Convolutional Neural Networks (CNNs). Our study aims to highlight the relevance of domain-related features for the extraction of the information of interest thanks to their joint nonlinear [...] Read more.

In this work, we face the problem of forest mapping from TanDEM-X data by means of Convolutional Neural Networks (CNNs). Our study aims to highlight the relevance of domain-related features for the extraction of the information of interest thanks to their joint nonlinear processing through CNN. In particular, we focus on the main InSAR features as the backscatter, coherence, and volume decorrelation, as well as the acquisition geometry through the local incidence angle. By using different state-of-the-art CNN architectures, our experiments consistently demonstrate the great potential of deep learning in data fusion for information extraction in the context of synthetic aperture radar signal processing and specifically for the task of forest mapping from TanDEM-X images. We compare three state-of-the-art CNN architectures, such as ResNet, DenseNet, and U-Net, obtaining a large performance gain over the baseline approach for all of them, with the U-Net solution being the most effective one. Full article

(This article belongs to the Special Issue SAR for Forest Mapping)

► Show Figures

Figure 1

32 pages, 14875 KB

Open AccessEditor’s ChoiceArticle

Overall Methodology Design for the United States National Land Cover Database 2016 Products

by Suming Jin, Collin Homer, Limin Yang, Patrick Danielson, Jon Dewitz, Congcong Li, Zhe Zhu, George Xian and Danny Howard

Remote Sens. 2019, 11(24), 2971; https://doi.org/10.3390/rs11242971 - 11 Dec 2019

Cited by 249 | Viewed by 15725

Abstract

The National Land Cover Database (NLCD) 2016 provides a suite of data products, including land cover and land cover change of the conterminous United States from 2001 to 2016, at two- to three-year intervals. The development of this product is part of an [...] Read more.

The National Land Cover Database (NLCD) 2016 provides a suite of data products, including land cover and land cover change of the conterminous United States from 2001 to 2016, at two- to three-year intervals. The development of this product is part of an effort to meet the growing demand for longer temporal duration and more frequent, accurate, and consistent land cover and change information. To accomplish this, we designed a new land cover strategy and developed comprehensive methods, models, and procedures for NLCD 2016 implementation. Major steps in the new procedures consist of data preparation, land cover change detection and classification, theme-based postprocessing, and final integration. Data preparation includes Landsat imagery selection, cloud detection, and cloud filling, as well as compilation and creation of more than 30 national-scale ancillary datasets. Land cover change detection includes single-date water and snow/ice detection algorithms and models, two-date multi-index integrated change detection models, and long-term multi-date change algorithms and models. The land cover classification includes seven-date training data creation and 14-run classifications. Pools of training data for change and no-change areas were created before classification based on integrated information from ancillary data, change-detection results, Landsat spectral and temporal information, and knowledge-based trajectory analysis. In postprocessing, comprehensive models for each land cover theme were developed in a hierarchical order to ensure the spatial and temporal coherence of land cover and land cover changes over 15 years. An initial accuracy assessment on four selected Landsat path/rows classified with this method indicates an overall accuracy of 82.0% at an Anderson Level II classification and 86.6% at the Anderson Level I classification after combining the primary and alternate reference labels. This methodology was used for the operational production of NLCD 2016 for the Conterminous United States, with final produced products available for free download. Full article

► Show Figures

Graphical abstract

24 pages, 3953 KB

Open AccessEditor’s ChoiceArticle

Evaluation of GPM-era Global Satellite Precipitation Products over Multiple Complex Terrain Regions

by Yagmur Derin, Emmanouil Anagnostou, Alexis Berne, Marco Borga, Brice Boudevillain, Wouter Buytaert, Che-Hao Chang, Haonan Chen, Guy Delrieu, Yung Chia Hsu, Waldo Lavado-Casimiro, Bastian Manz, Semu Moges, Efthymios I. Nikolopoulos, Dejene Sahlu, Franco Salerno, Juan-Pablo Rodríguez-Sánchez, Humberto J. Vergara and Koray K. Yilmaz

Remote Sens. 2019, 11(24), 2936; https://doi.org/10.3390/rs11242936 - 7 Dec 2019

Cited by 93 | Viewed by 7678

Abstract

The great success of the Tropical Rainfall Measuring Mission (TRMM) and its successor Global Precipitation Measurement (GPM) has accelerated the development of global high-resolution satellite-based precipitation products (SPP). However, the quantitative accuracy of SPPs has to be evaluated before using these datasets in [...] Read more.

The great success of the Tropical Rainfall Measuring Mission (TRMM) and its successor Global Precipitation Measurement (GPM) has accelerated the development of global high-resolution satellite-based precipitation products (SPP). However, the quantitative accuracy of SPPs has to be evaluated before using these datasets in water resource applications. This study evaluates the following GPM-era and TRMM-era SPPs based on two years (2014–2015) of reference daily precipitation data from rain gauge networks in ten mountainous regions: Integrated Multi-SatellitE Retrievals for GPM (IMERG, version 05B and version 06B), National Oceanic and Atmospheric Administration (NOAA)/Climate Prediction Center Morphing Method (CMORPH), Global Satellite Mapping of Precipitation (GSMaP), and Multi-Source Weighted-Ensemble Precipitation (MSWEP), which represents a global precipitation data-blending product. The evaluation is performed at daily and annual temporal scales, and at 0.1 deg grid resolution. It is shown that GSMaPV07 surpass the performance of IMERGV06B Final for almost all regions in terms of systematic and random error metrics. The new orographic rainfall classification in the GSMaPV07 algorithm is able to improve the detection of orographic rainfall, the rainfall amounts, and error metrics. Moreover, IMERGV05B showed significantly better performance, capturing the lighter and heavier precipitation values compared to IMERGV06B for almost all regions due to changes conducted to the morphing, where motion vectors are derived using total column water vapor for IMERGV06B. Full article

(This article belongs to the Special Issue Precipitation and Water Cycle Measurements Using Remote Sensing)

► Show Figures

Figure 1

14 pages, 9476 KB

Open AccessEditor’s ChoiceArticle

Use of Remote Sensing, Geophysical Techniques and Archaeological Excavations to Define the Roman Amphitheater of Torreparedones (Córdoba, Spain)

by Antonio Monterroso-Checa, Teresa Teixidó, Massimo Gasparini, José Antonio Peña, Santiago Rodero, Juan Carlos Moreno and José Antonio Morena

Remote Sens. 2019, 11(24), 2937; https://doi.org/10.3390/rs11242937 - 7 Dec 2019

Cited by 19 | Viewed by 6474

Abstract

Non-destructive techniques are widely used to explore and detect burial remains in archaeological sites. In this study, we present two sets of sensors, aerial and geophysics, that we have combined to analyze a 2 ha sector of ground in the Torreparedones Archaeological Park [...] Read more.

Non-destructive techniques are widely used to explore and detect burial remains in archaeological sites. In this study, we present two sets of sensors, aerial and geophysics, that we have combined to analyze a 2 ha sector of ground in the Torreparedones Archaeological Park located in Cordoba, Spain. Aerial platforms were used in a first step to identify a Roman amphitheater located near the Roman city. To ensure greater reliability and to rule out geological causes, a geophysical survey was subsequently carried out. Magnetic gradiometer, electrical resistivity tomography (ERT), and ground-penetrating radar (GPR) methods were also used to confirm the existence of this structure, define the geometry and, to the greatest possible extent, determine the degree of preservation of this construction. The adverse conditions for data acquisition was one of the main constraints, since the area of interest was an almond plantation which conditioned geophysical profiles. In addition, due to the low dielectric and magnetic contrast between the structures and the embedding material, meticulous data processing was required. In order to obtain further evidence of this amphitheater and to corroborate the aerial images and the geophysical models, an archaeological excavation was carried out. The results confirmed the cross-validation with the predicted non-destructive models. Therefore, this work can serve as an example to be used prior to conservation actions to investigate the suburbs and landscapes near similar roman cities in Spain. Full article

► Show Figures

Graphical abstract

32 pages, 16817 KB

Open AccessEditor’s ChoiceArticle

Evaluation of UAV LiDAR for Mapping Coastal Environments

by Yi-Chun Lin, Yi-Ting Cheng, Tian Zhou, Radhika Ravi, Seyyed Meghdad Hasheminasab, John Evan Flatt, Cary Troy and Ayman Habib

Remote Sens. 2019, 11(24), 2893; https://doi.org/10.3390/rs11242893 - 4 Dec 2019

Cited by 143 | Viewed by 16736

Abstract

Unmanned Aerial Vehicle (UAV)-based remote sensing techniques have demonstrated great potential for monitoring rapid shoreline changes. With image-based approaches utilizing Structure from Motion (SfM), high-resolution Digital Surface Models (DSM), and orthophotos can be generated efficiently using UAV imagery. However, image-based mapping yields relatively [...] Read more.

Unmanned Aerial Vehicle (UAV)-based remote sensing techniques have demonstrated great potential for monitoring rapid shoreline changes. With image-based approaches utilizing Structure from Motion (SfM), high-resolution Digital Surface Models (DSM), and orthophotos can be generated efficiently using UAV imagery. However, image-based mapping yields relatively poor results in low textured areas as compared to those from LiDAR. This study demonstrates the applicability of UAV LiDAR for mapping coastal environments. A custom-built UAV-based mobile mapping system is used to simultaneously collect LiDAR and imagery data. The quality of LiDAR, as well as image-based point clouds, are investigated and compared over different geomorphic environments in terms of their point density, relative and absolute accuracy, and area coverage. The results suggest that both UAV LiDAR and image-based techniques provide high-resolution and high-quality topographic data, and the point clouds generated by both techniques are compatible within a 5 to 10 cm range. UAV LiDAR has a clear advantage in terms of large and uniform ground coverage over different geomorphic environments, higher point density, and ability to penetrate through vegetation to capture points below the canopy. Furthermore, UAV LiDAR-based data acquisitions are assessed for their applicability in monitoring shoreline changes over two actively eroding sandy beaches along southern Lake Michigan, Dune Acres, and Beverly Shores, through repeated field surveys. The results indicate a considerable volume loss and ridge point retreat over an extended period of one year (May 2018 to May 2019) as well as a short storm-induced period of one month (November 2018 to December 2018). The foredune ridge recession ranges from 0 m to 9 m. The average volume loss at Dune Acres is 18.2 cubic meters per meter and 12.2 cubic meters per meter within the one-year period and storm-induced period, respectively, highlighting the importance of episodic events in coastline changes. The average volume loss at Beverly Shores is 2.8 cubic meters per meter and 2.6 cubic meters per meter within the survey period and storm-induced period, respectively. Full article

► Show Figures

Graphical abstract

18 pages, 8924 KB

Open AccessEditor’s ChoiceArticle

Next Generation Mapping: Combining Deep Learning, Cloud Computing, and Big Remote Sensing Data

by Leandro Parente, Evandro Taquary, Ana Paula Silva, Carlos Souza and Laerte Ferreira

Remote Sens. 2019, 11(23), 2881; https://doi.org/10.3390/rs11232881 - 3 Dec 2019

Cited by 65 | Viewed by 12322

Abstract

The rapid growth of satellites orbiting the planet is generating massive amounts of data for Earth science applications. Concurrently, state-of-the-art deep-learning-based algorithms and cloud computing infrastructure have become available with a great potential to revolutionize the image processing of satellite remote sensing. Within [...] Read more.

The rapid growth of satellites orbiting the planet is generating massive amounts of data for Earth science applications. Concurrently, state-of-the-art deep-learning-based algorithms and cloud computing infrastructure have become available with a great potential to revolutionize the image processing of satellite remote sensing. Within this context, this study evaluated, based on thousands of PlanetScope images obtained over a 12-month period, the performance of three machine learning approaches (random forest, long short-term memory-LSTM, and U-Net). We applied these approaches to mapped pasturelands in a Central Brazil region. The deep learning algorithms were implemented using TensorFlow, while the random forest utilized the Google Earth Engine platform. The accuracy assessment presented F1 scores for U-Net, LSTM, and random forest of, respectively, 96.94%, 98.83%, and 95.53% in the validation data, and 94.06%, 87.97%, and 82.57% in the test data, indicating a better classification efficiency using the deep learning approaches. Although the use of deep learning algorithms depends on a high investment in calibration samples and the generalization of these methods requires further investigations, our results suggest that the neural network architectures developed in this study can be used to map large geographic regions that consider a wide variety of satellite data (e.g., PlanetScope, Sentinel-2, Landsat-8). Full article

► Show Figures

Graphical abstract

21 pages, 6101 KB

Open AccessEditor’s ChoiceArticle

Interpretation of ASCAT Radar Scatterometer Observations Over Land: A Case Study Over Southwestern France

by Daniel Chiyeka Shamambo, Bertrand Bonan, Jean-Christophe Calvet, Clément Albergel and Sebastian Hahn

Remote Sens. 2019, 11(23), 2842; https://doi.org/10.3390/rs11232842 - 29 Nov 2019

Cited by 21 | Viewed by 4877

Abstract

This paper investigates to what extent soil moisture and vegetation density information can be extracted from the Advanced Scatterometer (ASCAT) satellite-derived radar backscatter (σ°) in a data assimilation context. The impact of independent estimates of the surface soil moisture (SSM) and [...] Read more.

This paper investigates to what extent soil moisture and vegetation density information can be extracted from the Advanced Scatterometer (ASCAT) satellite-derived radar backscatter (σ°) in a data assimilation context. The impact of independent estimates of the surface soil moisture (SSM) and leaf area index (LAI) of diverse vegetation types on ASCAT σ° observations is simulated over southwestern France using the water cloud model (WCM). The LAI and SSM variables used by the WCM are derived from satellite observations and from the Interactions between Soil, Biosphere, and Atmosphere (ISBA) land surface model, respectively. They permit the calibration of the four parameters of the WCM describing static soil and vegetation characteristics. A seasonal analysis of the model scores shows that the WCM has shortcomings over karstic areas and wheat croplands. In the studied area, the Klaus windstorm in January 2009 damaged a large fraction of the Landes forest. The ability of the WCM to represent the impact of Klaus and to simulate ASCAT σ° observations in contrasting land-cover conditions is explored. The difference in σ° observations between the forest zone affected by the storm and the bordering agricultural areas presents a marked seasonality before the storm. The difference is small in the springtime (from March to May) and large in the autumn (September to November) and wintertime (December to February). After the storm, hardly any seasonality was observed over four years. This study shows that the WCM is able to simulate this extreme event. It is concluded that the WCM could be used as an observation operator for the assimilation of ASCAT σ° observations into the ISBA land surface model. Full article

(This article belongs to the Special Issue Soil Moisture Retrieval using Radar Remote Sensing Sensors)

► Show Figures

Graphical abstract

Journal Menu

Journal Browser

Editor’s Choice Articles

Further Information

Guidelines

MDPI Initiatives

Follow MDPI