remotesensing-logo

Journal Browser

Journal Browser

Deforestation Detection with Deep Learning from Multispectral and Hyperspectral Satellite Images

A special issue of Remote Sensing (ISSN 2072-4292). This special issue belongs to the section "Forest Remote Sensing".

Deadline for manuscript submissions: closed (20 February 2024) | Viewed by 8523

Special Issue Editors

State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China
Interests: solar radiation; remote sensing; deforestation detection; regional climate analysis
State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences (CAS), Beijing 100101, China
Interests: remote sensing; vegetation recovery; surface solar radiation; cloud motion; land cover changes; climate impacts
Special Issues, Collections and Topics in MDPI journals

E-Mail Website
Guest Editor
Department of Geography, Environment, and Sustainability, The University of North Carolina at Greensboro, Greensboro, NC 27412, USA
Interests: spectral unmixing analysis; environmental planning; land use and land cover change modelling
Special Issues, Collections and Topics in MDPI journals

Special Issue Information

Dear Colleagues,

Deforestation is one of today's most serious global environmental problems, and its detrimental effects on climate regulation, ecosystem services, carbon storage, biodiversity, terrain stability, water retention, and economic development have been widely reported. With regard to carbon emissions in particular, studies have shown that the destruction of forests is one of the world's largest sources of greenhouse gas emissions, second only to the burning of fossil fuels. Consequently, deforestation is of special concern in the context of the global flight against climate warming and achieving carbon reduction.

To monitor the spatial and temporal evolution of deforestation, suitable detection methods must be developed based on emerging technologies such as remote sensing and deep learning. Traditional object-based algorithms provide promising results, but they require expertise to design task-specific features and cannot be applied at large scales. Deep-learning-based methods can capture the spatial and spectral patterns of multispectral and hyperspectral satellite images, bringing along with them novel ways for the addressal of urgent challenges in deforestation detection. This Special Issue aims to collect and share the progress in algorithm design and regional applications of deep learning for the detection of deforestation from remote sensing images.

Articles covering, but not limited to, the following topics are welcome in this Special Issue:

  • Deep learning algorithms for deforestation detection;
  • Applications of multispectral and hyperspectral satellite images for deforestation detection;
  • Datasets on the distribution and evolution of global or regional deforestation;
  • Intelligent labelling methods for deforestation detection and segmentation;
  • Analysis of climate effects associated with deforestation based on remote sensing;
  • Projections of future deforestation changes under carbon neutrality actions.

Dr. Ning Lu
Dr. Le Yu
Dr. Hou Jiang
Dr. Wenliang Li
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Remote Sensing is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2700 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • deforestation detection
  • environmental protection
  • remote sensing
  • deep learning
  • semantic segmentation
  • climate change
  • multispectral and hyperspectral images
  • convergence research

Published Papers (5 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Research

20 pages, 7584 KiB  
Article
A Learning Strategy for Amazon Deforestation Estimations Using Multi-Modal Satellite Imagery
by Dongoo Lee and Yeonju Choi
Remote Sens. 2023, 15(21), 5167; https://doi.org/10.3390/rs15215167 - 29 Oct 2023
Viewed by 1789
Abstract
Estimations of deforestation are crucial as increased levels of deforestation induce serious environmental problems. However, it is challenging to perform investigations over extensive areas, such as the Amazon rainforest, due to the vast size of the region and the difficulty of direct human [...] Read more.
Estimations of deforestation are crucial as increased levels of deforestation induce serious environmental problems. However, it is challenging to perform investigations over extensive areas, such as the Amazon rainforest, due to the vast size of the region and the difficulty of direct human access. Satellite imagery can be used as an effective solution to this problem; combining optical images with synthetic aperture radar (SAR) images enables deforestation monitoring over large areas irrespective of weather conditions. In this study, we propose a learning strategy for multi-modal deforestation estimations on this basis. Images from three different satellites, Sentinel-1, Sentinel-2, and Landsat 8, were utilized to this end. The proposed algorithm overcomes visibility limitations due to a long rainy season of the Amazon by creating a multi-modal dataset using supplementary SAR images, achieving high estimation accuracy. The dataset is composed of satellite data taken on a daily basis with relatively less monthly generated, ground truth masking data, which is called the many-to-one-mask condition. The Normalized Difference Vegetation Index and Normalized Difference Soil Index bands are selected to comprise the datasets. This yields better detection performance and a shorter training time than datasets consisting of RGB or all bands. Multiple deep neural networks are independently trained for each modality and an appropriate fusion method is developed to detect deforestation. The proposed method utilizes the distance similarity of the predicted deforestation rate to filter prediction results. The elements with high degrees of similarity are merged into the final result with average and denoising operations. The performances of five network variants of the U-Net family are compared, with Attention U-Net observed to exhibit the best prediction results. Finally, the proposed method is utilized to estimate the deforestation status of novel queries with high accuracy. Full article
Show Figures

Figure 1

23 pages, 19803 KiB  
Article
Spatio-Temporal Characteristics of Ice–Snow Freezing and Its Impact on Subtropical Forest Fires in China
by Xuecheng Wang, Xing Gao, Yuming Wu, Hou Jiang and Peng Wang
Remote Sens. 2023, 15(21), 5118; https://doi.org/10.3390/rs15215118 - 26 Oct 2023
Viewed by 953
Abstract
Ice–snow freezing may disrupt the growth condition and structure of forest vegetation, increasing combustible loads and thus triggering forest fires. China’s subtropical regions are rich in forest resources, but are often disturbed by ice–snow freezing, especially due to climate change. Clarifying the responsive [...] Read more.
Ice–snow freezing may disrupt the growth condition and structure of forest vegetation, increasing combustible loads and thus triggering forest fires. China’s subtropical regions are rich in forest resources, but are often disturbed by ice–snow freezing, especially due to climate change. Clarifying the responsive areas and times of forest fires to ice-snow freezing in this region is of vital importance for local forest fire management. In this study, meteorological data from 2001 to 2019 were used to extract the precipitation and its duration during the freezing period in order to analyze the freezing condition of forest vegetation in subtropical China. To improve the accuracy of identifying forest fires, we extracted forest fire information year-by-year and month-by-month based on the moderate resolution imaging spectroradiometer (MODIS) active fire data (MOD14A2) using the enhanced vegetation index (EVI), and analyzed the forest fire clustering characteristics in the region using the Moran’s Index. Then, correlation analysis between forest fires and freezing precipitation was utilized to explore the responsive areas and periods of forest fires caused by ice–snow freezing. Our analysis shows the following: (1) during the period of 2001–2019, the ice–snow freezing of forest vegetation was more serious in Hunan, Jiangxi, Hubei, and Anhui provinces; (2) forest fires in subtropical China have shown a significant downward trend since 2008 and their degree of clustering has been reduced from 0.44 to 0.29; (3) forest fires in Hunan, Jiangxi, and Fujian provinces are greatly affected by ice–snow freezing, and their correlation coefficients are as high as 0.25, 0.25, and 0.32, respectively; and (4) heavy ice–snow freezing can increase forest combustibles and affect forest fire behavior in February and March. This research is valuable for forest fire management in subtropical China and could also provide a reference for other regions. Full article
Show Figures

Graphical abstract

23 pages, 14733 KiB  
Article
Coupled MOP and PLUS-SA Model Research on Land Use Scenario Simulations in Zhengzhou Metropolitan Area, Central China
by Pengfei Guo, Haiying Wang, Fen Qin, Changhong Miao and Fangfang Zhang
Remote Sens. 2023, 15(15), 3762; https://doi.org/10.3390/rs15153762 - 28 Jul 2023
Cited by 5 | Viewed by 1228
Abstract
Land use simulations are critical in predicting the impact of land use change (LUC) on the Earth. Various assumptions and policies influence land use structure and are a key factor in decisions made by policymakers. Meanwhile, the spatial autocorrelation effect between land use [...] Read more.
Land use simulations are critical in predicting the impact of land use change (LUC) on the Earth. Various assumptions and policies influence land use structure and are a key factor in decisions made by policymakers. Meanwhile, the spatial autocorrelation effect between land use types has rarely been considered in existing land use spatial simulation models, and the simulation accuracy needs to be further improved. Thus, in this study, the driving mechanisms of LUC are analyzed. The quantity demand and spatial distribution of land use are predicted under natural development (ND), economic development (ED), ecological protection (EP), and sustainability development (SD) scenarios in Zhengzhou based on the coupled Multi-Objective Programming (MOP) model and the Patch-generating Land Use Simulation model (PLUS) considering Spatial Autocorrelation (PLUS-SA). We conclude the following. (1) The land use type in Zhengzhou was mainly cultivated land, and 83.85% of the land for urban expansion was cultivated land from 2000 to 2020. The reduction in forest from 2010 to 2020 was less than that from 2000 to 2010 due to the implementation of the policy in which farmland is transformed back into forests. (2) The accuracy of PLUS-SA was better than that of the traditional PLUS and Future Land Use Simulation (FLUS) models, and its Kappa coefficient, overall accuracy, and FOM were 0.91, 0.95, and 0.29, respectively. (3) Natural factors (temperature, precipitation, and DEM) contributed significantly to the expansion of cultivated land, and the increase in forest, grass, and construction land was greatly affected by socioeconomic factors (population, GDP, and proximity to town). (4) The land use structure will be more in line with the current requirements for sustainable urban development in the SD scenario, and the economic and ecological benefits will increase by 0.75 × 104 billion CNY and 1.71 billion CNY, respectively, in 2035 compared with those in 2020. The PLUS-SA model we proposed had higher simulation accuracy in Zhengzhou Compared with the traditional PLUS and FLUS models, and our research framework can provide a basis for decision-makers to formulate sustainable land use development policies to achieve high-quality and sustainable urban development. Full article
Show Figures

Figure 1

23 pages, 8656 KiB  
Article
A Combined Deep Learning and Prior Knowledge Constraint Approach for Large-Scale Forest Disturbance Detection Using Time Series Remote Sensing Data
by Bing Du, Zhanliang Yuan, Yanchen Bo and Yusha Zhang
Remote Sens. 2023, 15(12), 2963; https://doi.org/10.3390/rs15122963 - 7 Jun 2023
Cited by 2 | Viewed by 1431
Abstract
The scale and severity of forest disturbances across the globe are increasing due to climate change and human activities. Remote sensing analysis using time series data is a powerful approach for detecting large-scale forest disturbances and describing detailed forest dynamics. Various large-scale forest [...] Read more.
The scale and severity of forest disturbances across the globe are increasing due to climate change and human activities. Remote sensing analysis using time series data is a powerful approach for detecting large-scale forest disturbances and describing detailed forest dynamics. Various large-scale forest disturbance detection algorithms have been proposed, but most of them are only suitable for detecting high-magnitude forest disturbances (e.g., fire, harvest). Conversely, more continuous, subtle, and gradual lower-magnitude forest disturbances (e.g., thinning, pests, and diseases) have been subject to less focus. Deep learning (DL) can distinguish subtle differences in information within time series data, offering new opportunities to capture forest disturbances in a complete and detailed way. This study proposes an approach for analyzing forest dynamics across large areas and long time periods by combining DL time series classification and prior knowledge constraint. The approach consists of two stages: (1) an improved self-attention model used for time series classification to identify sequences with forest disturbance characteristics; (2) developed skip-disturbance recovery index (S-DRI) characterizing the temporal context, using prior knowledge constraint to identify forest disturbance years in time series with disturbance characteristics. In this study, the year of forest disturbances in five study areas located in the United States, Canada, and Poland from 2001 to 2020 was mapped. A total of 3082 manually interpreted test data with different disturbance causal agents (such as fire, harvest, conversion, hurricane, and pests) were sampled from five research areas for validation. Our approach was also evaluated against two forest disturbance benchmark datasets derived from LandTrendr and the Global Forest Change (GFC) dataset. The results demonstrate that our approach achieved an overall accuracy of 87.8%, surpassing the accuracy of LandTrendr (84.6%) and the Global Forest Change dataset (81.4%). Furthermore, our approach demonstrated lower omission rates (ranging from 10.0% to 67.4%) in detecting subtle to severe causal agents of forest disturbance, in comparison to LandTrendr (with a range of 18.0% to 81.6%) and GFC (with a range of 15.0% to 88.8%). This study, which involved mapping large-scale and long-term forest disturbance in multiple regions, revealed that our approach can be applied to new areas without a requirement for complex parameter adjustments. These results demonstrate the potential of our approach in generating comprehensive and detailed forest disturbance data, thus providing a new and effective method in this domain. Full article
Show Figures

Figure 1

17 pages, 4663 KiB  
Article
Predicting Spatially Explicit Composite Burn Index (CBI) from Different Spectral Indices Derived from Sentinel 2A: A Case of Study in Tunisia
by Mouna Amroussia, Olga Viedma, Hammadi Achour and Chaabane Abbes
Remote Sens. 2023, 15(2), 335; https://doi.org/10.3390/rs15020335 - 5 Jan 2023
Cited by 2 | Viewed by 1938
Abstract
Fire severity, which quantifies the degree of organic matter consumption, is an important component of the fire regime. High-severity fires have major ecological implications, affecting carbon uptake, storage and emissions, soil nutrients, and plant regeneration, among other ecosystem services. Accordingly, spatially explicit maps [...] Read more.
Fire severity, which quantifies the degree of organic matter consumption, is an important component of the fire regime. High-severity fires have major ecological implications, affecting carbon uptake, storage and emissions, soil nutrients, and plant regeneration, among other ecosystem services. Accordingly, spatially explicit maps of the fire severity are required to develop improved tools to manage and restore the most damaged areas. The aim of this study is to develop spatially explicit maps of the field-based fire severity (composite burn index—CBI) from different spectral indices derived from Sentinel 2A images and using several regression models. The study areas are two recent large fires that occurred in Tunisia in the summer of 2021. We employed different spectral severity indices derived from the normalized burn ratio (NBR): differenced NBR (dNBR), relative differenced NBR (RdNBR), and relativized burn Ratio (RBR). In addition, we calculated the burned area index for Sentinel 2 (BAIS2) and the thermal anomaly index (TAI). Different tree decision models (i.e., the recursive partitioning regression method [RPART], bagging regression trees [Bagging], and boosted regression trees [BRT]), as well as a generalized additive model [GAM]), were applied to predict the CBI. The main results indicated that RBR, followed by dNBR, were the most important spectral severity indices for predicting the field-based CBI. Moreover, BRT was the best regression model, explaining 92% of the CBI variance using the training set of points and 88% when using the validation set. These results suggested the adequacy of RBR index derived from Sentinel 2A for assessing and mapping forest fire severity in Mediterranean forests. These spatially explicit maps of field-based CBI could help improve post-fire recovery and restoration efforts. Full article
Show Figures

Figure 1

Back to TopTop