Next Article in Journal
Development of a New Analytical Method for the Characterization and Quantification of the Organic and Inorganic Carbonaceous Fractions in Snow Samples Using TOC and TOT Analysis
Next Article in Special Issue
Machine Learning Model-Based Estimation of XCO2 with High Spatiotemporal Resolution in China
Previous Article in Journal
Frost Mitigation Techniques for Tunnels in Cold Regions: The State of the Art and Perspectives
Previous Article in Special Issue
A Numerically Sensitive Study of Two Continuous Heavy-Pollution Episodes in the Southern Sichuan Basin of China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Atmosphere
Volume 14
Issue 2
10.3390/atmos14020370
atmosphere-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Uncertainty Analysis of Remote Sensing Underlying Surface in Land–Atmosphere Interaction Simulated Using Land Surface Models

by
Xiaolu Ling
1,2,*,
Hao Gao
3,4,
Jian Gao
1,
Wenhao Liu
1 and
Zeyu Tang
1
1
School of Environment and Spatial Informatics, China University of Mining and Technology, Xuzhou 221116, China
2
Jiangsu Key Laboratory of Coal-Based Greenhouse Gas Control and Utilization, China University of Mining and Technology, Xuzhou 221008, China
3
Key Laboratory of Radiometric Calibration and Validation for Environmental Satellites, National Satellite Meteorological Center (National Center for Space Weather), China Meteorological Administration, Beijing 100081, China
4
Innovation Center for FengYun Meteorological Satellite (FYSIC), Beijing 100081, China
*
Author to whom correspondence should be addressed.
Atmosphere 2023, 14(2), 370; https://doi.org/10.3390/atmos14020370
Submission received: 16 December 2022 / Revised: 6 February 2023 / Accepted: 7 February 2023 / Published: 13 February 2023
(This article belongs to the Special Issue Frontiers in Atmospheric Remote Sensing and Modelling)
Browse Figures
Versions Notes

Abstract

:
This paper reports a comparative experiment using remote sensing underlying surface data (ESACCI) and Community Land Model underlying surface data (CLM_LS) to analyze the uncertainty of land surface types in land–atmosphere interaction. The results showed that the global distribution of ESACCI cropland is larger than that of CLM_LS, and there is a great degree of difference in some regions, which can reach more than 50% regionally. Furthermore, the changes of the underlying surface conditions can be transmitted to the model results through the data itself, resulting in the uncertainty of the surface energy balance, surface micro-meteorological elements, and surface water balance simulated by the model, which further affects the climate simulation effect.

1. Introduction

Land–atmosphere interaction can affect the distribution of energy and matter in the earth system and can further influence atmospheric movement and, thus, climate change at different tempo-spatial scales (for example, from 0.1 m to global spatial scales and from 0.1 s to interannual or even longer time scales) [1,2]. Underlying surface types, such as forest, grassland, cropland, urban areas, and water, have different thermal and dynamic characteristics of the land surface that can have a great impact on climate at micro, local, and even global scales [3,4,5]. The crop type is one of the underlying surfaces that is greatly affected by human activities, and interventions and forces on nature can become more profound and extensive through the transmission of land surface processes [6,7,8,9].
Changes in the underlying surface have brought different degrees of climatic effects. Research results generally show that changes in vegetation cover can affect radiation and water balance by changing surface properties, such as surface albedo, roughness, soil moisture, and so on. Furthermore, changes of land cover can ultimately influence regional climate change, such as atmospheric temperature, humidity, precipitation, circulation, and so on [10,11]. Other results show that land surface processes have no obvious impact on climate at a global scale, but there are large differences at regional scales [9,12]. Different types of underlays have a significant impact on the diurnal variation of meteorological elements, and the degree of influence for the wind speed is the largest, followed by land surface temperature and air temperature, and air humidity has the lowest influence [3,5,13]. In terms of climatic effects, woodland has climatic effects of cooling, humidification, and wind protection; grassland has climatic effects of cooling and humidification; bare land is characterized by dryness and heat; and the climate effects for cropland is between bare land and grassland [14,15].
The underlying surface is affected by both natural factors and human activities. From the perspective of natural factors, different solar radiation, sea, and land conditions, etc., and their relative climate changes will result in different underlying surfaces and variation in characteristics [16]. Human activities have greatly altered morphological characteristics of the land surface, and this effect has been increasing [17,18,19,20]. Human activities have greatly altered land use and land cover, primarily in the past few centuries, by converting natural ecosystems to cropland [21,22]. Deforestation, urbanization, overgrazing, and desertification have been found combined with drought, extreme high temperature, and other climatic conditions. The IPCC Fifth Assessment Report mentioned that changes in land surface albedo brought about by land use and land cover change (LULCC) resulted in a change of −0.15 ± 0.10 Wm−2 since 1750 in the global radiative albedo [23]. In addition to radiative effects, changes due to LULCC also have large effects on nonradiative forcing and evapotranspiration, especially at regional scales [24]. To correctly assess the impact of climate change, it is necessary to accurately describe the conditions of the land surface.
Farming is an important component of typical human activities that affect climate [16,25,26]. Since the 21st century, the increase in the underlying surface of cropland has been very significant. For example, the world’s cropland has grown rapidly in the past three hundred years, from 2.65 × 108 hm2 in 1700 to 14.71 × 108 hm2 in 1990 [27]. According to FAO data, by 2009, the cropland area reached 15.33 × 108 hm2 [28]. There has been an increase in the underlying surface of cropland in most regions of the world, mainly in North America, Europe, the Middle East, India, East Asia, and Southeast Asia. Among them, the proportion of the underlying surface of cropland in some regions has increased by more than 50%. Most studies have found that climate change has a significant impact on cropland ecosystems; on the contrary, cropland ecosystems also have a moderating effect on climate change. For example, 10–15% of the increase in methane is attributable to paddy fields [29,30,31].
With the development of technology and remote sensing, there has been great progress in research on land surface processes, for which land surface data have great scientific significance. On the one hand, research on land surface processes demands extensive data; on the other hand, multi-parameter high-quality data also play a role in promoting research on land surface processes [32,33].
In the study of land surface processes, a variety of data products have emerged from multiple sources, such as the European Space Agency products (ESACCI) and GLOBCOVER land cover data products, which are mainly from MERIS FR/SR [34], MODIS land cover data product (MCD12Q1), etc. Although derived from satellite inversion data, these multi-source-based key parameter data products have differences in characteristics such as resolution and inversion methods, which, in turn, lead to differences among the data at a global scale. Large differences exist for various remotely sensed products to describe land surface features, which creates uncertainty in the land surface data itself. Using multi-source land cover data, Meiyappan and Jain [35] compared different estimated cropland and pasture areas and concluded that different estimation methods have an impact on the amount of cropland and pasture area, and it is likely that the uncertainty of the data will bring about uncertainty of the model.
Uncertainty in the data is transmitted to the model results through model inputs, which affects the climate simulations. Some studies used potential vegetation and existing vegetation as different model parameters to input into the Community Land Model (CLM) and obtained the change response of different parameters, as well as the climate impact brought about by LULCC [21,36]. Madhusoodhanan et al. [37] assessed uncertainty in global land cover products applied to hydroclimate models, to evaluate the uncertainty of simulation effects for evapotranspiration, latent heat flux, and sensible heat flux.
In this paper, the remotely sensed underlying surface data are substituted into the model as the underlying surface parameters to analyze the radiation, energy balance, and material exchange processes in the land–atmosphere interaction, as well as the ground energy balance state, and then analyze the impact of climate uncertainty by underlying data at the global and regional scales.

2. Data and Methods

2.1. Community Land Model (CLM4.5) and Its Underlying Surface Dataset

The model used in this paper is the Community Land Model version 4.5 (CLM4.5) developed and maintained by the National Center for Atmospheric Research (NCAR), which is widely used to simulate physical, chemical, and biological processes of terrestrial ecosystems interacting with the atmosphere at different tempo-spatial scales. Each modeling grid point contains five land cover types: glaciers, lakes, wetlands, cities, and vegetation, and vegetation is further divided into 15 plant functional types (PFTs). During simulation, CLM4.5 calculates the energy and water cycle at each PFTs separately and finally presents them at grid scale. It is worth mentioning that during the operation of CLM4.5, the dynamic carbon and nitrogen cycle module is also activated (CLM4.5CN), and the interaction of vegetation carbon–nitrogen elements can be considered in the process of model operation.
The underlying surface data of CLM4.5CN (CLM_LS) come from different databases (Olesen et al., 2013), among which the PFTs data come from the MODIS data [38], which are from a multi-year averaged underlying surface dataset, while cropland data are from [39].

2.2. ESACCI Underlying Surface Dataset

The ESACCI underlying surface data come from the combined data product of the MERIS satellite data and the SPOT-VGT 300 m spatial resolution underlying surface distribution. Starting from the data of different resolution channels of the MERIS satellite, the time series was obtained through preprocessing and then supplemented by the vegetation time series data of SPOT-VGT, plus the classification of land cover types, and finally the global land cover map was obtained. ESA CCI released a long-term global land cover time series (1992–2018) until now.

2.3. Processing of ESACCI Underlying Dataset

Table 1 shows the comparison of classification for CLM_LS and ESACCI, respectively. It can be seen that although both classifications divide the underlying surface into ice and snow, forest, grassland, shrub, cropland, water body, cities, and other types, the specific classification standards are inconsistent. In order to make the underlying surface of ESACCI the same as the CLM_LS PFTs, the ESACCI 300 m resolution land cover data is processed by up-grid [40]. The ESACCI land surface dataset is reclassified to the CLM land use categories according to the classification types as showed in the left column of Table 2, to make both underlying datasets consistent with each other. Furthermore, relative influence factors, such as accumulated temperature, growing degree days, precipitation, and other meteorological factors, are considered to divide the climatic zones into tropical, temperate, and boreal. The specific classification basis and classification method are shown in Table 2. For example, the regions with accumulated temperature larger than 15.5 °C are treated as tropical regions, while the regions with accumulated temperature less than −19 °C are treated as boreal regions, otherwise the regions are treated as temperate regions. As for the shrub and tree, the regions with growing degree days larger than 1200 days are treated as temperate regions, while less than 1200 days are treated as boreal regions. For the type of grass, the regions with growing degree days of less than 1000 days are treated as arctic regions, while values larger than 1000 days are treated as C3 and C4 grass regions, and the winter temperature and monthly precipitation are used to distinguish the C3 and C4 type. The detailed description for specific classification basis and classification method are shown in Table 2. In this manuscript we choose the year 2002 as the studying period, as shown in Table 3.

2.4. Experimental Design

Two comparative experiments (CTL and ESACCI) are designed by running the CLM4.5CN mode. The initial conditions, as well as the forcing conditions of both experiments, are consistent with each other. Only the underlying surface uses CLM_LS and ESACCI, respectively. After running the model, the global distribution and zonal distribution of key parameters, such as energy balance, micro-meteorological elements, and water balance, are discussed for analysis.

3. Results and Discussion

3.1. Differences in Underlying Surfaces

Figure 1 shows the spatial distribution of the difference in grid proportion for all PFTs in the CTL and ESACCI experiment, respectively. Generally, the global distribution characteristics of the 15 PFTs for both datasets remain consistent. Specifically, obviously negative biases between ESACCI and CLM_LS are found for NET-boreal and BDS–boreal, but positive biases for NET–temperate, NDT–boreal, and Crop. Large biases are more easily detected at high-latitude regions of the northern hemisphere, as well as for the vegetation with a needle leaf. The crop type is another typical type with a large difference, which is highly correlated with human activities. The reason for the difference may be that after preprocessing under different classification methods, there is an increase in the proportion of the data itself. Another possible reason for the change in the ratio is that the vegetation in high latitudes is ignored to a certain extent in the data observation process.
Specifically, the proportion of NET–boreal for ESACCI is very low, while for CLM_LS, most of the high latitudes of the northern hemisphere can reach more than 50%. The same condition occurs for the BDS–boreal. BDS–boreal has a large-scale and high-density distribution in the high latitudes of the northern hemisphere, reaching more than 50% in eastern Europe. However, in the global coverage of ESACCI, there is no distribution or sparse distribution for BDS–boreal.
The distribution of the Crop type described by CLM_LS and ESACCI is mainly concentrated in Europe, southern North America, the eastern coastal area of South America, central Africa, southeastern Africa, as well as East and Southeast Asia, India, and western Russia. The regions with the most intensive distribution of crop are mainly in India and central North America, where the proportion of cropland components can reach about 80%. Although the spatial distribution of cropland for both datasets is similar, there are still significant differences at global and regional scales. The proportion of crops in the global ESACCI data is larger than that in the CLM_LS data globally. It can be seen that in the Mediterranean coastal area of Europe, the proportion of cropland of CLM_LS is larger than that of ESACCI, and the difference is 20~30%, while in the eastern coastal area of India, the proportion of cropland of CLM_LS is also larger, with the difference reaching 40% to 50%. In other regions of the Eurasian continent, ESACCI’s cropland proportion is larger than that of CLM_LS, among which northwestern India, western Russia, northern China, and northeastern China are the most significant, reaching more than 50%.
The global proportions for all PFTs, as well as simulated variables in different latitudinal bands, are also discussed in this work based on climatic background divisions. The four bands are the northern and southern equatorial zones (0–23° N and 0–23° S), the northern temperate zones (23–45° N), and the boreal zone (45–65° N). As for the zonal range, the largest inconsistency between ESACCI and CLM_LS is found for the boreal zone, especially for the type of BDT-tropical, C3 grass, C4 grass, and crop. Compared with CLM_LS, most PFTs are underestimated, except for the fact that crops are overestimated globally and at four latitudinal bands. There is no difference for NDT–boreal between ESACCI and CLM_LS, mainly because the vegetation type is sparsely distributed globally (Figure 2).
In conclusion, there are various degrees of differences in the underlying surfaces between multi-source data sets, which come from observation methods, data generation process chains, and resolution classification methods that are preserved after preprocessing. Uncertainty is directly reflected at the global and regional scales for the underlying surface types.

3.2. Radiation

In order to analyze the impact of the underlying surface dataset on the surface energy balance, this section analyzes the spatial distribution of the key variables in the surface energy balance, including absorbed solar radiation (ASR), reflected solar radiation (RSR), and net longwave radiation (NLR). Considering that vegetation is mainly distributed in the northern hemisphere and July is the main growing season in the northern hemisphere, the month of July was selected as a typical month to analyze the global distribution of modeled variables.
At the global scale, the difference of ASR between ESACCI and CLM_LS is larger than NLR and RSR. There is an obvious band of negative bias, with a difference between −15 and −1 W/m2 for ASR at around 60° N latitude, that is, the ESACCI ASR is larger than CLM_LS ASR. In the southern part of North America, where cropland is widely distributed, there is essentially a positive difference, with the maximum reaching about 20 W/m2, located on the west coast of the United States at about 40° N latitude. On the Eurasian continent at mid-latitudes, regions of positive and negative biases are staggered. At the junction of Asia and Europe, where cropland is densely distributed, there is an obvious positive bias region, with a value ranging from 1 to 10 W/m2. The regions with the largest negative biases are located in the southwestern region of Russia, which happens to be one of the most densely distributed areas of cropland on the Eurasian continent and also one of the areas with the largest bias in the composition of cropland described by both datasets. The semi-arid regions of China with large positive differences in cropland components present a wide range of negative differences in ASR, ranging from 1 to 10 W/m2. India is also an important cropland distribution region in Asia, with a difference in regional composition of about 30%. Corresponding to Figure 3a, there is an insignificant negative difference distribution in the Indian region, with a value between −10 and −5 W/m2 for the ASR.
In addition to North America and Eurasia, there are two other regions with both dense cropland and large biases; one is the southeastern region of South America, and the other is the belt-like area around 10° N in central Africa, with the differences in cropland components all above 50%. In Figure 3a, there are positive and negative differences for ASR in these two typical areas, which are not the expected areas with a single negative difference, but some of the negative biases can reach 15 W/m2. There are also two areas with large values; one is the west coast of Africa around 10° S, with a positive difference in ASR reaching more than 20 W/m2, and the other is Australia, with a negative difference in ASR reaching 15 W/m2. However, the differences in the composition of the cropland corresponding to these two regions are very small, which indicates that there is a complex relationship between the difference in ASR and the difference in the composition of the cropland. Furthermore, the influence of some other vegetation types is another important factor.
In the surface energy balance, the change in the albedo of the land surface leads to a change in the ratio of ASR and RSR, which, in turn, affects the energy balance state of the land surface. Figure 3b shows the spatial distribution of differences for RSR. Combining Figure 3a,b, it can be found that there is an obvious reverse distribution at the global scale; that is, the regions with positive biases in ASR will have negative biases in RSR. In regions where cropland is concentrated, such as semi-arid regions of China, India, southeastern North America, central Africa, and southeastern South America, RSR has the opposite distribution of characteristics to ASR. Globally, RSR shows the largest negative bias of −15~−10 W/m2 in southern Africa, and the west coast of the United States in North America. In the semi-arid region of China and northeastern China, there is an obviously positive bias.
There is a large-scale negatively biased region for NLR at about 60° N latitudes, with the maximum value of −16~−12 W/m2 occurring in Vladivostok, Russia. On the contrary, at the mid-high latitude (50~60° N) region near the Sino-Mongolian border, a positive bias of about 20 W/m2 exists. The two significant regions with positive and negative NLR biases correspond to insignificant differences in cropland composition. There is no obvious difference for NLR in the cropland-intensive areas of the Eurasian continent (semi-arid regions of China, India, the Middle East, and Europe). Similar to what is found in Figure 3a, there is a maximum negative bias of NLR in Australia (about −16 W/m2), and a maximum value of positive bias of NLR (above 30 W/m2) in southern Africa. The biases for NLR in North America are also significant. Most areas of North America have different degrees of negative biases. The maximum negative bias of −20 W/m2 appears on the coast of the Arctic Ocean around 70° N, and there is also an obvious negative bias area in southeastern North America, where cropland is densely distributed (−16~−4 W/m2). At 50~60° N latitude, there is a certain degree of positive difference, and the largest value appears on the border of the United States and Canada, around 20 W/m2. In summary, although cropland composition causes general changes in global NLR, the distribution of NLR in Figure 3c is somewhat different to the distribution of the cropland component.
The zonal distribution of differences in ASR, RSR, and NLR between ESACCI and CLM_LS is also displayed in Figure 4. Generally, the ASR and RSR are obviously larger in the northern temperate region than in the other regions, especially in SON seasons. The difference in NLR is the largest in the boreal region, with the maximum in September. There is also a large bias for NLR in the northern temperate region, with the maximum in January.

3.3. Heat Fluxes

In addition to radiation balance, sensible heat fluxes (Hs) and latent heat fluxes (Le) also play an important role in the process of surface energy balance. Figure 5 shows the difference in spatial distribution of Hs and Le, respectively.
Figure 5a shows that the difference in global Hs is staggered. The largest positive bias for Hs is found in Africa around 10° S latitude, with values larger than 40 W/m2, which is also the center with the largest positive difference between ASR and NLR, and the center with the largest negative difference in RSR.
On the Eurasian continent, there is a significant negative bias region for Hs (with the largest bias of 15–20 W/m2) concentrated in western Russia and Europe where the cropland is densely distributed and the cropland composition varies greatly. There is a positive difference in the eastern part of Russia and the border area between China and Russia, with a positive difference center around 20 W/m2. In the vast cropland-intensive areas in East China and South China, as well as in India, there is a negative difference for Hs, ranging from 5 to 20 W/m2. In the southeastern part of North America with dense cropland, the distribution of Hs differences is similar to that of differences in cropland composition, with both positive and negative differences coexisting. The difference is that a negative difference center of about 20 W/m2 appears in the eastern part of the continent at 40–50° N latitude, while a positive difference center appears along the coast of the Arctic Ocean at about 60° N latitude. This distribution characteristic is similar to the distribution of ASR differences. There is also a negative difference area of about 10 W/m2 in the difference area of cropland components in central Africa and southeastern South America, and a large negative difference area of about 20 W/m2 in the central part of South America.
Figure 5b presents the global distribution of Le differences. It is obvious that the differences are mainly concentrated in the southeastern part of North America, most of Europe, the Indian region, and the southern part of eastern China, all of which are the regions with the densest distribution of cropland. There is a general positive difference in Le in the southeastern part of North America, with a maximum value of more than 16 W/m2. The positive difference areas in Europe are mainly distributed in Central Europe, Northern Europe, and the western part of Russia, and the differences in Le are in the range of 3–16 W/m2, while in the Mediterranean coastal area, there is a negative difference area as high as −12 W/m2. There is also a positive difference of 3~12 W/m2 for Le in the Indian region. There are large-scale areas of positive differences in Le in the eastern and southern parts of China, with the highest value of 16 W/m2 in the southeastern coastal area, which corresponds to the high-value area of positive difference in the southeastern coast of China in the difference map of cropland components.
Combining Figure 5 and Figure 6, it can be seen that there is a certain opposite distribution for the difference between Hs and Le, and the centers of their large values correspond to the areas with a dense distribution of cropland or areas with large differences in cropland components. It can be seen that the uncertainty of the multi-source cropland dataset, when brought into the model, has a large impact on the sensible and latent heat flux parts of the surface energy balance.

3.4. Response of Micro-Meteorological Elements

The most intuitive impact on the change in the lower boundary conditions in reality and in the model is the change in temperature and humidity. In this section, the distribution of the difference of 2 m air temperature (T2m) and relative humidity (RH) is utilized for analysis.
Figure 7 shows the regional distribution of T2m and RH differences, respectively, and the distributions of RH shows opposite changes to the T2m. Most of the Eurasian continent has a negative bias for T2m, and the value is within 0.6 °C. For the densely distributed cropland areas in the east and south of China, the negative bias is within 0.2 °C. In Mongolia, there is a clear positive difference of more than 1 °C. The negative difference center of T2m is found in the central part of Australia and the center of North America; the negative difference center in central Australia reaches more than 1.2 °C, while the negative difference center in North America is only about 1 °C. Conversely, there is a positive difference of 1.2°C in the southern region of Australia. Comparing the distribution of the differences in cropland components, these large-value centers are consistent with the areas with large-value differences in crop composition. For example, the area with a large negative difference of T2m in North America corresponds to a large negative difference in cropland components.
As for the RH, most of the Eurasian continent has a positive bias within 3%. In the densely distributed areas of cropland in the east and south of China, there is a positive bias within 1%. In Mongolia, there is a clear negative bias of about −4%. The positive difference center of RH appeared in the central part of Australia and the center of North America, and the positive bias between the two places reached more than 4%; the negative bias center appeared in the southern part of Africa and the western coastal area of South America, with a negative bias of 4%. Comparing the distribution of cropland shows that these large-value centers are consistent with the large-value areas of component differences. For example, a large positive difference for RH in North America corresponds to a large negative difference in cropland composition.
Combining Figure 7a,b, it can be seen that there is a good correlation between the distribution of differences in cropland components and the distribution of RH and T2m differences, but these relationships are only the results of qualitative analysis, which need to be confirmed by quantitative analysis results. Contrary to the other variables, T2m and RH showed larger bias during the JJA seasons for all bands, except that the T2m and RH are large in the MAM and SON seasons for the boreal regions.
The zonal distribution of differences in T2m and RH between ESACCI and CLM_LS are also displayed in Figure 8. Generally, the T2m and RH represent reverse changes for all zonal bands. The differences of land surface make the earth warmer and drier, except for the boreal region. In the boreal region, the T2m difference is larger during cold seasons. While for the other regions, the largest T2m difference is found in its warm seasons. The largest difference of RH is found in the Northern Equatorial region all through the year, and a larger difference of RH is also found in the boreal region during its cold seasons.

3.5. Response of Hydrological Variables

Hydrological processes are also an important part of the land–atmosphere interaction, so this paper also discusses hydrological variables closely related to vegetation, including canopy evaporation and transpiration.
Figure 9 shows the spatial distribution of the difference for canopy evaporation and transpiration, respectively. It can be seen that there is a good distribution consistency between canopy evaporation and transpiration differences, especially in the regions with a negative difference. The areas with a positive difference of canopy evaporation are mainly in most of Europe and the northern region of Russia, and the central and southern parts of North America, with the differences in the center all around 8 W/m2. In contrast, the areas of positive difference in canopy transpiration are mainly in the southeastern part of North America, the northern part of Eurasia, and East Asia and South Asia. The centers with large positive differences are located in the central and western regions of Russia and central North America (in good agreement with canopy evaporation), and the positive differences reach more than 20 W/m2.
Negative difference regions for canopy evaporation and transpiration vary. The negative difference area of canopy evaporation is widely distributed on the land at 30° S~30° N, among which 0°~10° N has a large belt-shaped area with a negative difference of canopy evaporation (about 6 W/m2). Another large value area is in Mongolia, where the negative difference of canopy evaporation can reach more than 8 W/m2. In contrast, the negative difference area of canopy transpiration is larger; for example, southern Africa, the eastern coast of South America, and northern North America all have large value areas of at least 12 W/m2. In addition, compared with the negative difference in canopy evaporation, the negative difference in canopy transpiration in Mongolia has a wider range, and the negative difference is more significant (above 20 W/m2).
Comprehensively analyzing canopy evaporation and transpiration, in combination with the difference in the distribution of cropland components, the different distribution of cropland and hydrological variables in East Asia, Europe, and most of the middle and low latitudes have a good fit with each other.
The zonal distribution of differences in evaporation and transpiration between ESACCI and CLM_LS are also displayed in Figure 10. Generally, both evaporation and transpiration have a large difference during cold seasons in the northern hemisphere. Negative differences are found for evaporation and transpiration all through the year, except for the boreal regions. The boreal evaporation has positive differences all through the year, especially in its cold seasons.
When the land surface changes, the first effect is canopy density that affects the solar radiation reaching the ground, as well as long-wave radiation and net radiation simultaneously. With the change of energy entered into the land surface, the distribution to sensible heat flux and latent heat flux will also change, as a result influence the temperature and relative humidity. Furthermore, the change of canopy density will also influence stomatal conductivity, and finally influence evapotranspiration of vegetation.

4. Discussion and Conclusions

This paper studies the differences of absorbed solar radiation, reflected solar radiation, net longwave radiation, sensible/latent heat fluxes, relative humidity, and 2 m air temperature, as well as canopy evaporation and transpiration caused by different underlying surfaces in the Community Land Model, to analyze the impact of the uncertainty of remote sensing data on different climate variables. The main conclusions are as follows.
  • The proportion of cropland distribution in ESACCI data is essentially larger than that in CLM_LS data globally, and it can reach more than 50% locally.
  • The model concluded that the difference in parameters is related to the difference in cropland distribution, and the difference in absorbed solar radiation and sensible heat flux has an obvious response to the difference in cropland distribution.
  • Differences in the distribution of cropland bring about different degrees of changes in climate parameters. The uncertainty of the data can be transmitted to the model operation process through the data themselves, which brings uncertainty to the model results and affects the climate simulation.
Future work will mainly focus on (i) introducing more observational data, such as GlobCover dataset, MODIS dataset, and FAO dataset, etc., to analyze the impact of uncertainty of different datasets on model simulation results; (ii) extending the data period for analysis, that is, not only analyzing the impact of the uncertainty of the cropland data set itself on the model simulated climate, but also considering the impact of the uncertainty of LUCC on the uncertainty caused by models; and (iii) coupling with atmospheric models to analyze the feedback effect between underlying surface changes and climate change.

Author Contributions

Formal analysis and investigation, X.L. and H.G.; conceptualization, X.L.; methodology, X.L. and H.G.; writing—original draft preparation, X.L., J.G. and Z.T.; writing—review and editing, X.L., H.G., J.G., W.L. and Z.T. All authors have read and agreed to the published version of the manuscript.

Funding

This work was jointly funded by the National Key R&D Program of China (2017YFA0604304), the National Science Foundation of China (Grant No. 42075114, 41705101), the Priority Academic Program Development of Jiangsu Higher Education Institutions (140119001), and the General Project of Modern Agriculture from the Primary R&D Program of Xuzhou (KC21132).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The European Space Agency CCI land cover dataset are available in section “land cover” at https://www.esa-landcover-cci.org/ (accessed on 16 December 2022).

Acknowledgments

We would like to thank the editor and the anonymous reviewers for exceptionally thoughtful reviews and suggestions that greatly improved this paper.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Desai, A.R.; Paleri, S.; Mineau, J.; Kadum, H.; Wanner, L.; Mauder, M.; Butterworth, B.J.; Durden, D.J.; Metzger, S. Scaling land-atmosphere interactions: Special or fundamental? J. Geophys. Res. Biogeosci. 2022, 127, e2022JG007097. [Google Scholar] [CrossRef]
  2. Bonan, G.B. Sensitivity of a GCM simulation to inclusion of inland water surfaces. J. Clim. 1995, 8, 14. [Google Scholar] [CrossRef]
  3. Berg, A.; Findell, K.; Lintner, B.; Giannini, A.; Seneviratne, S.I.; van den Hurk, B.; Lorenz, R.; Pitman, A.; Hagemann, S.; Meier, A.; et al. Land-atmosphere feedbacks amplify aridity increase over land under global warming. Nat. Clim. Change 2016, 6, 869–874. [Google Scholar] [CrossRef]
  4. Santanello, J.A., Jr.; Dirmeyer, P.A.; Ferguson, C.R.; Findell, K.L.; Tawfik, A.B.; Berg, A.; Ek, M.; Gentine, P.; Guillod, B.P.; van Heerwaarden, C.; et al. Land-atmosphere interactions: The LoCo perspective. Bull. Am. Meteorol. Soc. 2018, 99, 1253–1272. [Google Scholar] [CrossRef]
  5. Green, J.K.; Seneviratne, S.I.; Berg, A.M.; Findell, K.L.; Hagemann, S.; Lawrence, D.M.; Gentine, P. Large influence of soil moisture on long-term terrestrial carbon uptake. Nature 2019, 565, 476–479. [Google Scholar] [CrossRef] [PubMed]
  6. Warrach-Sagi, K.; Ilgwersen, J.; Schwitalla, T.; Troost, C.; Aurbacher, J.; Jach, L.; Berger, T.; Streck, T.; Wulfmeyerl, V. Noah-MP With the generic crop growth model gecros in the WRF model: Effects of dynamic crop growth on land-atmosphere interaction. J. Geophys. Res. Atmos. 2022, 127, e2022JD036518. [Google Scholar] [CrossRef]
  7. Maruyama, A.; Kuwagata, T. Coupling land surface and crop growth models to estimate the effects of changes in the growing season on energy balance and water use of rice paddies. Agric. For. Meteorol. 2010, 150, 919–930, Erratum in Agric. For. Meteorol. 2021, 308–309, 108585. [Google Scholar] [CrossRef]
  8. Majumder, A.; Kingra, P.K.; Setia, R.; Singh, S.P.; Pateriya, B. Influence of land use/land cover changes on surface temperature and its effect on crop yield in different agro-climatic regions of Indian Punjab. Geocarto Int. 2020, 35, 663–686. [Google Scholar] [CrossRef]
  9. Liu, F.S.; Chen, Y.; Xiao, D.P.; Bai, H.Z.; Tao, F.L.; Ge, Q.S. Modeling crop growth and land surface energy fluxes in wheat-maize double cropping system in the North China Plain. Theor. Appl. Climatol. 2020, 142, 959–970. [Google Scholar] [CrossRef]
  10. Baker, J.C.A.; de Souza, D.C.; Kubota, P.Y.; Buermann, W.; Coelho, C.A.S.; Andrews, M.B.; Gloor, M.; Garcia-Carreras, L.; Figueroa, S.N.; Spracklen, D.V. An assessment of land-atmosphere interactions over South America using satellites, reanalysis, and two global climate models. J. Hydrometeorol. 2021, 22, 905–922. [Google Scholar] [CrossRef]
  11. Song, Y.Y.; Wei, J.F. Diurnal cycle of summer precipitation over the North China Plain and associated land-atmosphere interactions: Evaluation of ERA5 and MERRA-2. Int. J. Climatol. 2021, 41, 6031–6046. [Google Scholar] [CrossRef]
  12. Ma, Y.M.; Hu, Z.Y.; Xie, Z.P.; Ma, W.Q.; Wang, B.B.; Chen, X.L.; Li, M.S.; Zhong, L.; Sun, F.L.; Gu, L.L.; et al. A long-term (2005–2016) dataset of hourly integrated land-atmosphere interaction observations on the Tibetan Plateau. Earth Syst. Sci. Data 2020, 12, 2937–2957. [Google Scholar] [CrossRef]
  13. Dare-Idowu, O.; Jarlan, L.; Le-Dantec, V.; Rivalland, V.; Ceschia, E.; Boone, A.; Brut, A. Hydrological Functioning of maize crops in Southwest France using eddy covariance measurements and a land surface model. Water 2021, 13, 1481. [Google Scholar] [CrossRef]
  14. Zhang, Z.; Chen, F.; Barlage, M.; Bortolotti, L.E.; Famiglietti, J.; Li, Z.; Ma, X.; Li, Y. Cooling effects revealed by modeling of wetlands and land-atmosphere interactions. Water Resour. Res. 2022, 58, e2021WR030573. [Google Scholar] [CrossRef]
  15. Imran, H.M.; Hossain, A.; Shammas, M.I.; Das, M.K.; Islam, M.R.; Rahman, K.; Almazroui, M. Land surface temperature and human thermal comfort responses to land use dynamics in Chittagong city of Bangladesh. Geomat. Nat. Haz. Risk 2022, 13, 2283–2312. [Google Scholar] [CrossRef]
  16. Mahmood, R.; Pielke, R.A.; Hubbard, K.G.; Niyogi, D.; Bonan, G.; Lawrence, P.; McNider, R.; McAlpine, C.; Etter, A.; Gameda, S.; et al. Impacts of land use/land cover change on climate and future research priorities. Bull. Am. Meteorol. Soc. 2010, 91, 37–46. [Google Scholar] [CrossRef]
  17. Zhai, J.; Liu, R.G.; Liu, J.Y.; Huang, L.; Qin, Y.W. Human-induced landcover changes drive a diminution of land surface albedo in the Loess Plateau (China). Remote Sens. 2015, 7, 2926–2941. [Google Scholar] [CrossRef]
  18. Fu, T.M.; Zhang, L.; Chen, B.W.; Yan, M. Human disturbance on the land surface environment in tropical islands: A remote sensing perspective. Remote Sens. 2022, 14, 2100. [Google Scholar] [CrossRef]
  19. Hu, S.L.; Yu, B.; Luo, S.; Zhuo, R.R. Spatial pattern of the effects of human activities on the land surface of China and their spatial relationship with the natural environment (s10668-021-01871-6, 2021). Environ. Dev. Sustain. 2022, 24, 14421. [Google Scholar] [CrossRef]
  20. Yamada, T.J.; Pokhrel, Y. Effect of Human-induced land disturbance on subseasonal predictability of near-surface variables using an atmospheric general circulation model. Atmosphere 2019, 10, 725. [Google Scholar] [CrossRef] [Green Version]
  21. Xu, Z.; Mahmood, R.; Yang, Z.-L.; Fu, C.; Su, H. Investigating diurnal and seasonal climatic response to land use and land cover change over monsoon Asia with the Community Earth System Model. J. Geophys. Res. Atmos. 2015, 120, 1137–1152. [Google Scholar] [CrossRef]
  22. Okkan, U.; Kirdemir, U. Investigation of the behavior of an agricultural-operated dam reservoir under RCP scenarios of AR5-IPCC. Water Resour. Manag. 2018, 32, 2847–2866. [Google Scholar] [CrossRef]
  23. Pachauri, R.K.; Allen, M.R.; Barros, V.R.; Broome, J.; Cramer, W.; Christ, R.; Church, J.A.; Clarke, L.; Dahe, Q.; Dasgupta, P.; et al. Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change; IPCC: Geneva, Switzerland, 2014; p. 151. [Google Scholar]
  24. Birch, E.L. Climate change 2014: Impacts, adaptation, and vulnerability. J. Am. Plann. Assoc. 2014, 80, 184–185. [Google Scholar] [CrossRef]
  25. Foley, J.A.; DeFries, R.; Asner, G.P.; Barford, C.; Bonan, G.; Carpenter, S.R.; Chapin, F.S.; Coe, M.T.; Daily, G.C.; Gibbs, H.K.; et al. Global consequences of land use. Science 2005, 309, 570–574. [Google Scholar] [CrossRef] [PubMed]
  26. Zhang, X.Z.; Wang, W.C.; Fang, X.Q.; Ye, Y.; Zheng, J.Y. Agriculture development-induced surface albedo changes and climatic implications across Northeastern China. Chin. Geogr. Sci. 2012, 22, 264–277. [Google Scholar] [CrossRef]
  27. Goldewijk, K.K. Estimating global land use change over the past 300 years: The HYDE database. Glob. Biogeochem. Cycles 2001, 15, 417–433. [Google Scholar] [CrossRef]
  28. Cao, F.Q.; Dan, L.; Ma, Z.G. Simulative study of the impact of the cropland change on the regional climate over China. Acta Meteorol. Sin. 2015, 73, 14. [Google Scholar]
  29. Ahmad, M.J.; Cho, G.H.; Kim, S.H.; Lee, S.; Adelodun, B.; Choi, K.S. Influence mechanism of climate change over crop growth and water demands for wheat-rice system of Punjab, Pakistan. J. Water Clim. Change 2021, 12, 1184–1202. [Google Scholar] [CrossRef]
  30. Aurbacher, J.; Parker, P.S.; Sanchez, G.A.C.; Steinbach, J.; Reinmuth, E.; Ingwersen, J.; Dabbert, S. Influence of climate change on short term management of field crops—A modelling approach. Agric. Syst. 2013, 119, 44–57. [Google Scholar] [CrossRef]
  31. Cheng, J.Q.; Yin, S.Y. Quantitative assessment of climate change impact and anthropogenic influence on crop production and food security in Shandong, Eastern China. Atmosphere 2022, 13, 1160. [Google Scholar] [CrossRef]
  32. Wu, G.X.; Lin, H.; Zou, X.L.; Liu, B.Q.; He, B. Research on global climate change and scientific data. Adv. Earth Sci. 2014, 29, 8. [Google Scholar]
  33. Fathololoumi, S.; Firozjaei, M.K.; Li, H.J.; Biswas, A. Surface biophysical features fusion in remote sensing for improving land crop/cover classification accuracy. Sci. Total Environ. 2022, 838, 156520. [Google Scholar] [CrossRef]
  34. Quaife, T.; Cripps, E. Bayesian analysis of uncertainty in the GlobCover 2009 land cover product at climate model grid scale. Remote Sens. 2016, 8, 314. [Google Scholar] [CrossRef]
  35. Meiyappan, P.; Jain, A.K. Three distinct global estimates of historical land-cover change and land-use conversions for over 200 years. Front. Earth Sci. 2012, 6, 122–139. [Google Scholar] [CrossRef]
  36. Lawrence, P.J.; Chase, T.N. Investigating the climate impacts of global land cover change in the community climate system model. Int. J. Climatol. 2010, 30, 2066–2087. [Google Scholar] [CrossRef]
  37. Madhusoodhanan, C.G.; Sreeja, K.G.; Eldho, T.I. Assessment of uncertainties in global land cover products for hydro-climate modeling in India. Water Resour. Res. 2017, 53, 1713–1734. [Google Scholar] [CrossRef]
  38. Lawrence, P.J.; Chase, T.N. Representing a new MODIS consistent land surface in the Community Land Model (CLM 3.0). J. Geophys. Res. Biogeosci. 2007, 112. [Google Scholar] [CrossRef]
  39. Ramankutty, N.; Foley, J.A. Estimating historical changes in land cover: North American croplands from 1850 to 1992. Glob. Ecol. Biogeogr. 1999, 8, 381–396. [Google Scholar] [CrossRef]
  40. Li, H.; Zhang, H.; Mamtimin, A.; Fan, S.; Ju, C. A new land-use dataset for the weather research and forecasting (WRF) model. Atmosphere 2020, 11, 350. [Google Scholar] [CrossRef] [Green Version]
Atmosphere 14 00370 g001 550
Figure 1. The difference in grid proportion for different PFTs (ESACCI–CLM_LS): (a) NET–temperate; (b) NET–boreal; (c) NDT–boreal; (d) BET–tropical; (e) BET–temperate; (f) BDT–tropical; (g) BDT–temperate; (h) BDT–boreal; (i) BES–temperate; (j) BDS–temperate; (k) BDS–boreal; (l) C3 arctic grass; (m) C3 grass; (n) C4 grass; (o) Crop.
Figure 1. The difference in grid proportion for different PFTs (ESACCI–CLM_LS): (a) NET–temperate; (b) NET–boreal; (c) NDT–boreal; (d) BET–tropical; (e) BET–temperate; (f) BDT–tropical; (g) BDT–temperate; (h) BDT–boreal; (i) BES–temperate; (j) BDS–temperate; (k) BDS–boreal; (l) C3 arctic grass; (m) C3 grass; (n) C4 grass; (o) Crop.
Atmosphere 14 00370 g001
Atmosphere 14 00370 g002 550
Figure 2. The difference in the PFTs’ coverage between ESACCI and CLM_LS.
Figure 2. The difference in the PFTs’ coverage between ESACCI and CLM_LS.
Atmosphere 14 00370 g002
Atmosphere 14 00370 g003 550
Figure 3. The difference of ESACCI and CLM_LS for (a) absorbed solar radiation (ASR), (b) reflected solar radiation (RSR), and (c) net longwave radiation (NLR) (unit: W/m2) (July 2000).
Figure 3. The difference of ESACCI and CLM_LS for (a) absorbed solar radiation (ASR), (b) reflected solar radiation (RSR), and (c) net longwave radiation (NLR) (unit: W/m2) (July 2000).
Atmosphere 14 00370 g003
Atmosphere 14 00370 g004 550
Figure 4. The zonal distribution of differences in (a) ASR, (b) RSR, and (c) NLR between ESACCI and CLM_LS (W/m2).
Figure 4. The zonal distribution of differences in (a) ASR, (b) RSR, and (c) NLR between ESACCI and CLM_LS (W/m2).
Atmosphere 14 00370 g004
Atmosphere 14 00370 g005 550
Figure 5. The difference of ESACCI and CLM_LS for (a) sensible heat fluxes (Hs) and (b) latent heat fluxes (Le) (W/m2) (July 2000).
Figure 5. The difference of ESACCI and CLM_LS for (a) sensible heat fluxes (Hs) and (b) latent heat fluxes (Le) (W/m2) (July 2000).
Atmosphere 14 00370 g005
Atmosphere 14 00370 g006 550
Figure 6. The zonal distribution of differences in (a) Hs and (b) Le between ESACCI and CLM_LS (W/m2).
Figure 6. The zonal distribution of differences in (a) Hs and (b) Le between ESACCI and CLM_LS (W/m2).
Atmosphere 14 00370 g006
Atmosphere 14 00370 g007 550
Figure 7. The difference of ESACCI and CLM_LS for (a) 2 m air temperature (T2m, unit: °C) and (b) relative humidity (RH, unit: %) (July 2000).
Figure 7. The difference of ESACCI and CLM_LS for (a) 2 m air temperature (T2m, unit: °C) and (b) relative humidity (RH, unit: %) (July 2000).
Atmosphere 14 00370 g007
Atmosphere 14 00370 g008 550
Figure 8. The zonal distribution of differences in (a) T2m (°C) and (b) RH (%) between ESACCI and CLM_LS.
Figure 8. The zonal distribution of differences in (a) T2m (°C) and (b) RH (%) between ESACCI and CLM_LS.
Atmosphere 14 00370 g008
Atmosphere 14 00370 g009 550
Figure 9. The difference between ESACCI and CLM_LS for (a) evaporation (unit: W/m2), and (b) transpiration (unit: W/m2) (July 2000).
Figure 9. The difference between ESACCI and CLM_LS for (a) evaporation (unit: W/m2), and (b) transpiration (unit: W/m2) (July 2000).
Atmosphere 14 00370 g009
Atmosphere 14 00370 g010 550
Figure 10. The zonal distribution of differences in (a) evaporation (W/m2) and (b) transpiration (W/m2) between ESACCI and CLM_LS.
Figure 10. The zonal distribution of differences in (a) evaporation (W/m2) and (b) transpiration (W/m2) between ESACCI and CLM_LS.
Atmosphere 14 00370 g010
Table
Table 1. Comparison of classification for CLM_LS and ESACCI.
Table 1. Comparison of classification for CLM_LS and ESACCI.
CLM_LSESACCI
1Needleleaf evergreen tree–temperate (NET–temperate)10Cropland, rainfed
2Needleleaf evergreen tree–boreal
(NET–boreal)		20Cropland, irrigated or post-flooding
3Needleleaf deciduous tree–boreal
(NDT–boreal)		30Mosaic cropland/natural vegetation
4Broadleaf evergreen tree–tropical
(BET–tropical)		40Mosaic natural vegetation/cropland
5Broadleaf evergreen tree–temperate
(BET–temperate)		50Tree cover, broadleaved, evergreen, closed to open
6Broadleaf deciduous tree–tropical
(BDT–tropical)		60Tree cover, broadleaved, deciduous, closed to open
7Broadleaf deciduous tree–temperate
(BDT–temperate)		70Tree cover,
Needleleaved, evergreen, closed to open
8Broadleaf deciduous tree–boreal
(BDT–boreal)		80Tree cover, needleleaved, deciduous, closed to open
9Broadleaf evergreen shrub–temperate (BES–temperate)90Tree cover,
mixed leaf type
10Broadleaf deciduous shrub–temperate (BDS–temperate)100Mosaic tree and shrub/herbaceous cover
11Broadleaf deciduous shrub–boreal (BDS–boreal)110Mosaic herbaceous cover/tree and shrub
12C3 arctic grass120Shrubland
13C3 grass130Grassland
14C4 grass140Lichens and mosses
15Crop150Sparse vegetation
0Bare soil160Tree cover, flooded, fresh or
brackish water
170Tree cover, flooded,
Saline water
180Shrub or herbaceous cover, flooded, fresh/saline/brackish water
190Urban areas
200Bare areas
210Water bodies
220Permanent snow and ice
0No Data
Table
Table 2. Classification basis and classification method.
Table 2. Classification basis and classification method.
No.PFTsESACCIClimate Rules
1NET–temperate70Tc1 > −19 °C, and GDD2 > 1200
2NET–boreal70Tc ≤ −19 °C, or GDD ≤ 1200
3NDT–boreal80, 81, 82none
4BET–tropical50Tc > 15.5 °C
5BET–temperate50Tc ≤ 15.5 °C
6BDT–tropical60, 61Tc > 15.5 °C
7BDT–temperate60, 61−15 °C < Tc ≤ 15.5 °C, and GDD > 1200
8BDT–boreal60, 61Tc ≤ −15 °C, or GDD ≤ 1200
9BES–temperate11, 120,
121, 122		Tc > −19 °C and GDD > 1200 and Pann3 > 520 nm and Pwin4 > 2/3 Pann
10BDS–temperate11, 120,
121,122		Tc > −19 °C and GDD > 1200 and (Pann ≤ 520 nm or Pwin ≤ 2/3 Pann)
11BDS–boreal11, 120,
121, 122		Tc ≤ −19 °C, or GDD ≤ 1200
12C3 arctic grass110, 130GDD < 1000
13C3 grass110, 130GDD > 1000 and (Tw5 ≤ 22 °C or Pmon6 ≤ 25 mm and for months with T > 22 °C)
14C4 grass110, 130GDD > 1000 and (Tw > 22 °C and driest month Pmon > 25 mm)
15Crop10, 20, 30None
1 Tc: accumulated temperature; 2 GDD: growing degree days; 3 Pann: annual precipitation; 4 Pwin: winter precipitation; 5 Tw: winter temperature; 6 Pmon: monthly precipitation.
Table
Table 3. Experimental design.
Table 3. Experimental design.
NameModelLand Surface TypesInitial
Condition		Spatial
Resolution		Simulation
PERIODS
CTLCLM4.5CNCLM_LSCLM0.9 × 1.252000
ESACCICLM4.5CNESACCICLM0.9 × 1.252000
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Ling, X.; Gao, H.; Gao, J.; Liu, W.; Tang, Z. Uncertainty Analysis of Remote Sensing Underlying Surface in Land–Atmosphere Interaction Simulated Using Land Surface Models. Atmosphere 2023, 14, 370. https://doi.org/10.3390/atmos14020370

AMA Style

Ling X, Gao H, Gao J, Liu W, Tang Z. Uncertainty Analysis of Remote Sensing Underlying Surface in Land–Atmosphere Interaction Simulated Using Land Surface Models. Atmosphere. 2023; 14(2):370. https://doi.org/10.3390/atmos14020370

Chicago/Turabian Style

Ling, Xiaolu, Hao Gao, Jian Gao, Wenhao Liu, and Zeyu Tang. 2023. "Uncertainty Analysis of Remote Sensing Underlying Surface in Land–Atmosphere Interaction Simulated Using Land Surface Models" Atmosphere 14, no. 2: 370. https://doi.org/10.3390/atmos14020370

APA Style

Ling, X., Gao, H., Gao, J., Liu, W., & Tang, Z. (2023). Uncertainty Analysis of Remote Sensing Underlying Surface in Land–Atmosphere Interaction Simulated Using Land Surface Models. Atmosphere, 14(2), 370. https://doi.org/10.3390/atmos14020370

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop