Land Surface Reflectance Research Articles

Comparison of different methods of determining land surface reflectance for AOD retrieval

Determining land surface reflectance (LSR) is important for satellite retrieval of aerosol optical depth (AOD). Many studies have focused on the development of AOD retrieval through using different land surface reflectance determination algorithms. However, there is still lack of a comprehensive comparison among those algorithms. In this study, AOD retrieval algorithms with different land surface reflectance assumptions were compared and analyzed using Advanced Himawari Imager (AHI) data. The results demonstrate that the AOD retrieved by the algorithm considering the land surface bidirectional reflectance distribution function (BRDF) effects (Method 3) has a higher accuracy than that based on the dark target (Method 1) and second lowest reflectance (Method 2) methods. The overall correlation coefficient between the AOD retrievals considering land surface BRDF and AERONET AOD measurements is 0.93 and the Root Mean Square Error (RMSE) is 0.13. In addition, there are 68.91 % of AOD matchups falling within the expected error (±0.05 ± 0.2 × AOD AERONET). Method 3 shows a more robust performance than the other two methods for different seasons, normalized difference vegetation index (NDVI) values and scattering angles. It also shows the advantages of AOD retrieval in urban regions with relatively high LSR. The AOD retrievals based on Method 2 show the best agreement with MODIS 1 km AOD product, while Method 1 AODs are much closer to the MODIS 10 km AODs. The differences of annual average values of AHI AOD retrievals using different algorithms can be high to 0.13 at 11:00 (Beijing Time) over the Beijing-Tianjin-Hebei (BTH) region, which could cause a 23.6Wm-2 discrepancy when calculating the direct aerosol radiative effect (DARE) at the bottom of the atmosphere.

DeepYield: A combined convolutional neural network with long short-term memory for crop yield forecasting

Crop yield forecasting is of great importance to crop market planning, crop insurance, harvest management, and optimal nutrient management. Commonly used approaches for crop prediction include but are not limited to conducting extensive manual surveys or using data from remote sensing. Considering the increasing amount of data provided by remote sensing imagery, this approach is becoming increasingly important for the task of crop yield forecasting and there is a need for more sophisticated approaches to extract the inherent spatiotemporal patterns of these data. Although considerable progress has been made in this field by using Deep Learning (DL) methods such as Convolutional Neural Networks (CNN), no study before has investigated the use of Convolutional Long Short-Term Memory (ConvLSTM) for crop yield forecasting. Here, we propose DeepYield, a combined structure, that integrates the ConvLSTM layers with the 3-Dimensional CNN (3DCNN) for more accurate and reliable spatiotemporal feature extraction. The models are trained by using county-based historical yield data and MODIS Land Surface Temperature (LST), Surface Reflectance (SR), and Land Cover (LC) data over 1836 primary soybean growing counites in the Contiguous United States (CONUS). The forecasting performance of the developed models is compared against the competing approaches including Decision Trees, CNN + GP, and CNN-LSTM and results indicate that DeepYield significantly outperforms these techniques and also performs better than both ConvLSTM and 3DCNN.

High resolution aerosol optical depth retrieval over urban areas from Landsat-8 OLI images

The satellite-retrieved aerosol optical depth (AOD) provides unique estimation of aerosol loading across a continuous space. However, current AOD products with a relatively coarse resolution (≥1 km) can hardly capture the details in urban areas with large spatial AOD gradients. To address this issue, here we developed a novel retrieval algorithm for retrieving AOD with extra fine spatial resolution (30 m) from Landsat-8 satellite OLI images. In the algorithm, the three land surface reflectance (LSR) estimation schemes and four aerosol types are adopted to improve the retrieval accuracy. The algorithm is applied on Beijing and Wuhan city in China during 2014–2019. Results suggest that the retrieved AOD with the new algorithm exhibits good agreement with the ground-based measured AOD (R2 = 0.920), and 81.63% of the AODs fall within the expected error line with a root-mean-square error of 0.112. Moreover, the high-resolution AOD products also successfully identified polluted sources and discrepancy of aerosol loading over different land cover types in two megacities in China, implying its great potential on detecting fine aerosol emission sources over the complex urban area. Such improvement of the new algorithm for retrieving 30 m AOD developed in this study demonstrates its substantial potentials in supporting further studies of air pollution management and human exposure at extra-fine spatial scale.

Vegetation Phenology in Permafrost Regions of Northeastern China Based on MODIS and Solar-induced Chlorophyll Fluorescence

Vegetation phenology is an indicator of vegetation response to natural environmental changes and is of great significance for the study of global climate change and its impact on terrestrial ecosystems. The normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI), extracted from the Moderate Resolution Imaging Spectrometer (MODIS), are widely used to monitor phenology by calculating land surface reflectance. However, the applicability of the vegetation index based on ‘greenness’ to monitor photosynthetic activity is hindered by poor observation conditions (e.g., ground shadows, snow, and clouds). Recently, satellite measurements of solar-induced chlorophyll fluorescence (SIF) from OCO-2 sensors have shown great potential for studying vegetation phenology. Here, we tested the feasibility of SIF in extracting phenological metrics in permafrost regions of the northeastern China, exploring the characteristics of SIF in the study of vegetation phenology and the differences between NDVI and EVI. The results show that NDVI has obvious SOS advance and EOS lag, and EVI is closer to SIF. The growing season length based on SIF is often the shortest, while it can represent the true phenology of vegetation because it is closely related to photosynthesis. SIF is more sensitive than the traditional remote sensing indices in monitoring seasonal changes in vegetation phenology and can compensate for the shortcomings of traditional vegetation indices. We also used the time series data of MODIS NDVI and EVI to extract phenological metrics in different permafrost regions. The results show that the length of growing season of vegetation in predominantly continuous permafrost (zone I) is longer than in permafrost with isolated taliks (zone II). Our results have certain significance for understanding the response of ecosystems in cold regions to global climate change.

Spatiotemporal Remote Sensing Image Fusion Using Multiscale Two-Stream Convolutional Neural Networks

Spatiotemporal remote sensing image fusion (STF) is a promising way to obtain remote sensing data with both fine spatial and temporal resolutions. Gradual and abrupt changes in land surface reflectance images are the main challenges in existing STF methods. Advanced deep learning techniques present powerful ability in learning image-changed information. Therefore, this article proposes a novel spatiotemporal image fusion method using multiscale two-stream convolutional neural networks (STFMCNNs). Multiscale two-stream convolutional neural networks are proposed to capture different sizes of objects in feature learning from a coarse spatial resolution image and two pairs of coarse and fine spatial resolution (FR) images at other dates. Meanwhile, temporal dependence and temporal consistency are explored as complementary information for STFMCNN. Moreover, a local fusion method is developed to characterize local variation by combining two predicted images derived from each stream. Two experiments on different real images are conducted to demonstrate the effectiveness of STFMCNN. Results show that STFMCNN outperformed three existing methods by predicting more accurate FR images with more preserved changed information.

Correction of Directional Effects in VEGETATION NDVI Time-Series

Land surface reflectance measurements from the VEGETATION program (SPOT-4, SPOT-5 and PROBA-V satellites) have led to the acquisition of consistent time-series of Normalized Difference Vegetation Index (NDVI) at a global scale. The wide imaging swath (>2000 km) of the family of VEGETATION space-borne sensors ensures a daily coverage of the Earth at the expense of a varying observation and illumination geometries between successive orbit overpasses for a given target located on the ground. Such angular variations infer saw-like patterns on time-series of reflectance and NDVI. The presence of directional effects is not a real issue provided that they can be properly removed, which supposes an appropriate BRDF (Bidirectional Reflectance Distribution Function) sampling as offered by the VEGETATION program. An anisotropy correction supports a better analysis of the temporal shapes and spatial patterns of land surface reflectance values and vegetation indices such as NDVI. Herein we describe a BRDF correction methodology, for the purpose of the Copernicus Global Land Service framework, which includes notably an adaptive data accumulation window and provides uncertainties associated with the NDVI computed with normalized reflectance. Assessing the general performance of the methodology in comparing time-series between normalized and directional NDVI reveals a significant removal of the high-frequency noise due to directional effects. The proposed methodology is computationally efficient to operate at a global scale to deliver a BRDF-corrected NDVI product based on long-term Time-Series of VEGETATION sensor and its follow-on with the Copernicus Sentinel-3 satellite constellation.

Spatiotemporal Evolution of Fractional Vegetation Cover and Its Response to Climate Change Based on MODIS Data in the Subtropical Region of China

The subtropical vegetation plays an important role in maintaining the structure and function of global ecosystems, and its contribution to the global carbon balance are receiving increasing attention. The fractional vegetation cover (FVC) as an important indicator for monitoring environment change, is widely used to analyze the spatiotemporal pattern of regional and even global vegetation. China is an important distribution area of subtropical vegetation. Therefore, we first used the dimidiate pixel model to extract the subtropical FVC of China during 2001–2018 based on MODIS land surface reflectance data, and then used the linear regression analysis and the variation coefficient to explore its spatiotemporal variations characteristics. Finally, the partial correlation analysis and the partial derivative model were used to analyze the influences and contributions of climate factors on FVC, respectively. The results showed that (1) the subtropical FVC had obvious spatiotemporal heterogeneity; the FVC high-coverage and medium-coverage zones were concentratedly and their combined area accounted for more than 70% of the total study area. (2) The interannual variation in the average subtropical FVC from 2001 to 2018 showed a significant growth trend. (3) In 76.28% of the study area, the regional FVC showed an increasing trend, and the remaining regional FVC showed a decreasing trend. However, the overall fluctuations in the FVC (increasing or decreasing) in the region were relatively stable. (4) The influences of climate factors to the FVC exhibited obvious spatial differences. More than half of all pixels exhibited the influence of the average annual minimum temperature and the annual precipitation had positive on FVC, while the average annual maximum temperature had negative on FVC. (5) The contributions of climate changes to FVC had obvious heterogeneity, and the average annual minimum temperature was the main contribution factor affecting the dynamic variations of FVC.

Land use and Land Cover Characteristics using Bhuvan and MODIS Satellite Data

Understanding vegetation characteristics is essential for watershed modeling, like in the prediction of streamflow and evapotranspiration (AET) estimation. So, the present study was taken to analyze the Land use/Land cover characteristics in a Sub-humid tropical river basin which is originating in the forested part of Western Ghats mountain ranges using the Moderate Resolution Imaging Spectroradiometer (MODIS) and Bhuvan satellite data as inputs for the algorithm. All the fourteen LU/LC characteristics present in the Hemavathi basin (5427 km2) were analyzed in the basin using satellite data which is located in Karnataka, India. Land Surface Reflectance (LSR) and Land Surface Temperature (LST) were the two data products used as input to map the pixel-wise variations in albedo, the fraction of vegetation (FV) and Land Surface Temperature (LST). It was found from the rainfall data that the year 2019 experienced higher rainfall than the average and 2012 very low rainfall than the normal. Parameters considered in this study Albedo, LST and FV are susceptible to wetness and temperature conditions. Variations in albedo and LST were similar in that both values in the summer of 2019 and 2012 are high than winter due to the high temperature and less wetness from all the LU/LC classes. Similarly, FV variations show opposite trends that values in the summer of 2019 and 2012 are low than in winter, which is due to the high temperature and less wetness. The results and discussions show that significant realistic variations in albedo, LST and FV with respect to all LU/LC classes. All the LU/LC classes characteristics in this study show significant variations with respect to wetness and temperature conditions, so the methodology proposed in this study can be used in regional monitoring of LU/LC classes in a convenient and cost-effective manner.

An Improved DDV Algorithm for the Retrieval of Aerosol Optical Depth From NOAA/AVHRR Data

Aerosol Optical Depth (AOD) is one of the important parameters to characterize the physical properties of the atmospheric aerosol, which is used to describe the extinction characteristics of the aerosol, and also to calculate the aerosol content, to assess the degree of air pollution and to study aerosol climate effect. To study the historical change of aerosol in long-time series, the advanced very high resolution radiometer (AVHRR) data earliest used for aerosol research was used in this study. Due to the lack of shortwave infrared (SWIR) (center at 2.13 µm) of the sensor, the relationship between the blue and red bands with SWIR cannot be provided, and the visible band used to calculate the normalized difference vegetation index (NDVI) contains the wavelength range of red and green, it is very difficult to calculate the accurate land surface reflectance (LSR). Therefore, based on the Dense Dark Vegetation algorithm (DDV), we propose to introduce mature MODIS vegetation index products (MYD13) to correct AVHRR NDVI, to support the estimation of AVHRR LSR, determine the relationship between corrected AVHRR NDVI and visible band LSR, and to carry out aerosol retrieval. The results show that about 63% of the data are within the error line, and there is a consistent distribution trend in the inter-comparison validation with MODIS aerosol products (MYD04).

An Improved Method for Retrieving Aerosol Optical Depth Using Gaofen-1 WFV Camera Data

The four wide-field-of-view (WFV) cameras aboard the GaoFen-1 (GF-1) satellite launched by China in April 2013 have been applied to the studies of the atmospheric environment. To highlight the advantages of GF-1 data in the atmospheric environment monitoring, an improved deep blue (DB) algorithm using only four bands (visible–near infrared) of GF-1/WFV was adopted to retrieve the aerosol optical depth (AOD) at ~500 m resolution in this paper. An optimal reflectivity technique (ORT) method was proposed to construct monthly land surface reflectance (LSR) dataset through converting from MODIS LSR product according to the WFV and MODIS spectral response functions to make the relationship more suitable for GF-1/WFV. There is a good spatial coincidence between our retrieved GF-1/WFV AOD results and MODIS/Terra or Himawari-8/AHI AOD products at 550 nm, but GF-1/WFV AOD with higher resolution can better characterized the details of regional pollution. Additionally, our retrieved GF-1/WFV AOD (2016–2019) results showed a good agreement with AERONET ground-based AOD measurements, especially, at low levels of AOD. Based on the same LSR dataset transmitted from 2016–2018 MODIS LSR products, RORT of 2016–2018 and 2019 GF-1/WFV AOD retrievals can reach up to 0.88 and 0.94, respectively, while both of RMSEORT are smaller than 0.13. It is indicated that using the ORT method to deal with LSR information can make GF-1/WFV AOD retrieval algorithm more suitable and flexible.

The Application of CYGNSS Data for Soil Moisture and Inundation Mapping in Australia

Cyclone global navigation satellite system (CYGNSS) has provided a valuable opportunity for high spatiotemporal monitoring of land surface reflectivity over the past few years. CYGNSS with a constellation of eight microsatellites is able to constantly observe the “scattered” global positioning system signals from the land. In this study, we validate the CYGNSS land reflectivity data in Australia for mapping the spatial extent of the inundated area and for determining temporal changes in surface soil moisture. CYGNSS level 1 data acquired for the period of 2017–2020 is assessed against various measurements, including satellite and ground-based measurements. Empirical mode decomposition is used to better analyze the CYGNSS time series and their relationship with the independent measurements. Furthermore, the mission's ability to capture surface reflectivity changes in response to extreme climatic events is analyzed. The results show that high spatial and temporal resolution CYGNSS data can largely represent the top layer ( $\sim$ 5 cm) soil moisture spatial and temporal variations close to soil moisture active passive. CYGNSS surface reflectivity results are also found to be sensitive to surface water changes and able to depict inundated land surface.

An LSWI-Based Method for Mapping Irrigated Areas in China Using Moderate-Resolution Satellite Data

Accurate spatial information about irrigation is crucial to a variety of applications, such as water resources management, water exchange between the land surface and atmosphere, climate change, hydrological cycle, food security, and agricultural planning. Our study proposes a new method for extracting cropland irrigation information using statistical data, mean annual precipitation and Moderate Resolution Imaging Spectroradiometer (MODIS) land cover type data and surface reflectance data. The approach is based on comparing the land surface water index (LSWI) of cropland pixels to that of adjacent forest pixels with similar normalized difference vegetation index (NDVI). In our study, we validated the approach over mainland China with 612 reference samples (231 irrigated and 381 non-irrigated) and found the accuracy of 62.09%. Validation with statistical data also showed that our method explained 86.67 and 58.87% of the spatial variation in irrigated area at the provincial and prefecture levels, respectively. We further compared our new map to existing datasets of FAO/UF, IWMI, Zhu and statistical data, and found a good agreement with the irrigated area distribution from Zhu’s dataset. Results show that our method is an effective method apply to mapping irrigated regions and monitoring their yearly changes. Because the method does not depend on training samples, it can be easily repeated to other regions.

Fire danger assessment using geospatial modelling in Mekong delta, Vietnam: Effects on wetland resources

The Mekong Delta (MD), Vietnam is a large tropical wetland important for commodity provision, water balance regulation, and biodiversity conservation. However, the delta is under severe fire risk due to prolonged drought and rising temperatures largely emanating from climatic change. Our study applied MODIS Terra land surface reflectance (MOD09A1), MODIS Terra land surface temperature (MOD11A2), and ASTER digital elevation model (DEM) to develop fire danger models to assess spatial patterns in fire danger in the MD. Four variables: Potential Surface Temperature (PST), Modified Normalized Difference Fire Index (MNDFI), Perpendicular Moisture Index (PMI) and Normalized Multiband Drought Index (NMDI), were extracted from the satellite products to develop the models using Analytical Hierarchical Process (AHP) and weighted overlay methods to generate fire danger maps during the 2019 fire season. The models were validated using MODIS Terra and Aqua thermal anomaly product (MOD14), with overall accuracy of 84.13%. The result demonstrated that the model accurately estimates fire risk over the MD. The study also found that forests, especially Melaleuca wetland forests in U Minh wetlands, and agricultural land were under severe fire risk. The implication is that measures such as proper management of forests and agricultural activities for mitigating forest fires and integrated fire and water strategies for either fire risk mitigation or biodiversity conservation are necessary for sustainable management of the MD wetland.

Modeling Bidirectional Polarization Distribution Function of Land Surfaces Using Machine Learning Techniques

Accurate estimation of polarized reflectance (Rp) of land surfaces is critical for remote sensing of aerosol optical properties. In the last two decades, many data-driven bidirectional polarization distribution function (BPDF) models have been proposed for accurate estimation of Rp, among which the generalized regression neural network (GRNN) based BPDF model has been reported to perform the best. GRNN is just a simple machine learning (ML) technique that can solve non-linear problems. Many ML techniques were reported to work well in solving non-linear problems and consequently may provide better performance in BPDF modeling. However, incorporating various ML techniques with BPDF modeling and comparing their performances have never been well documented. In this study, three widely used ML algorithms—i.e., support vector regression (SVR), K-nearest-neighbor (KNN), and random forest (RF)—were applied for BPDF modeling. Using measurements collected by the Polarization and Directionality of the Earth’s Reflectance onboard PARASOL satellite (POLDER/PARASOL), non-linear relationships between Rp and the input variables, i.e., Fresnel factor (Fp), scattering angle (SA), reflectance at 670 nm (R670) and 865 nm (R865), were built using these ML algorithms. Results showed that taking Fp, SA, R670, and R865 as input variables, the performance of the four ML-based BPDF models was quite similar. The KNN-based BPDF model provided slightly better results, and improved the accuracy of the semi-empirical BPDF models by 9.55% in terms of the overall root mean square error (RMSE). Experiments of different configuration of input variables suggested that using multi-band reflectance as input variables provided better results than using vegetation indices. The RF-based BPDF model using all reflectances at six bands as input variables produced the best results, improving the overall accuracy by 6.62% compared with the GRNN-based BPDF model. Among all the input variables, reflectance at absorbing spectral bands—e.g., 490 nm and 670 nm—played more significant roles in RF-based BPDF modeling due to the domination of polarized partition in total reflectance. Fresnel factor and scattering angle were also important for BPDF modeling. This study confirmed the feasibility of applying ML techniques to more accurate BPDF modeling, and the RF-based BPDF model proposed in this study can be used to increase the accuracy of remote sensing of the complete aerosol properties.

Multi-information collaborative cloud identification algorithm in Gaofen-5 Directional Polarimetric Camera imagery

The Directional Polarimetric Camera (DPC) onboard China's Gaofen-5 (GF-5) satellite has provided multi-angle (3-polarization channel), large-scale (1850 km swath width and 3.3 km spatial resolution), and high-frequency (2-day revisit period) Earth observation since May 2018. These features make DPC imagery application scenarios extensively. However, like other optical imagery, the presence of clouds is also a pervasive and unavoidable issue in DPC imagery. To leverage both radiation and polarization properties of DPC, we proposed a multi-information collaborative (MIC) method to identify clouds in the DPC imagery. Instead of a fixed single threshold, the MIC method adopts dynamic thresholds obtained by simulation in different atmosphere models, at different times, and under different underlying surfaces. Specifically, we included surface albedo and ice/snow cover distribution libraries to the MIC method, as they compensate for fewer spectral bands in the DPC imagery, thereby improving the accuracy of cloud detection results, especially in special bright surface scenarios (e.g., desert, bare soil and ice/snow). We also added an ice/snow detection algorithm to further eliminate the issue of misidentifying ice/snow pixels as clouds. Finally, after obtaining the DPC cloud mask results based on the MIC method, we calculated four cloud confidence levels for different application requirements by cloud quality evaluation criteria.We evaluated the MIC algorithm by comparing it with two other independent satellite cloud observation products, namely Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) and Moderate Resolution Imaging Spectroradiometer (MODIS). We found that the MIC cloud mask is in good agreement with the other two cloud products, with agreement probabilities of 93.06% (CALIPSO) and 85% (MODIS), respectively. Furthermore, the detected high confidence cloud and clear sky results agree with the CALIPSO cloud confidence products by more than 97.35% and 96.13%, respectively. We therefore suggest that the MIC method can provide the basis for subsequent studies of atmospheric parameters, such as accurate retrieval of aerosol optical thickness (AOT), cloud optical thickness (COT), cloud droplet effective radius (CDR) and land surface reflectance.

Reconstruction of Spatiotemporally Continuous MODIS-Band Reflectance in East and South Asia from 2012 to 2015

To reconstruct Moderate Resolution Imaging Spectroradiometer (MODIS) band reflectance with optimal spatiotemporal continuity, three bidirectional reflectance distribution function (BRDF) models—the Ross-Thick-Li-Sparse Reciprocal (RTLSR) model, Gao model, and adjusted BF model—were used to retrieve MODIS-band reflectance for cloudy MODIS pixels according to different inversion conditions with a proposed filling algorithm. Then, a spatiotemporally continuous MODIS-band reflectance dataset for most of Asia with more than 98% spatiotemporal coverage was reconstructed from 2012 to 2015. The validation highlighted an evident improvement in filling cloudy MODIS observations; a reasonable spatial distribution, such as in South Asia and Southeast Asia; and acceptable precision for the filled MODIS pixels, with the root mean square error percentage (RMSE%) at 9.7–9.8% and 12–16% for the Gao and adjusted BF models, respectively. In the course of reconstructing the spatiotemporal continuous MODIS-band reflectance, the differences among the three models were discussed further. For a 16-day period with a stable and unchanged land surface, the RTLSR model, as a basic model, accurately derived land surface reflectance (no more than 10% RMSE% for MCD43C1 V006 band 1) and outperformed the other two models. When the inversion period is sufficiently long (e.g., 108 days, 188 days, 268 days, or a full year), the Gao/adjusted BF model provides better precision than the RTLSR model by considering the normalized difference vegetation index (NDVI) and soil moisture/NDVI as intermediate variables used to adjust the BRDF parameters in real time. The Gao model is optimal when the inversion period is sufficiently long. Based on combining the RTLSR model and Gao/adjusted BF model, we proposed a filling algorithm to derive a dataset of MODIS-band reflectance with optimal spatiotemporal continuity.

An experimental sky-image-derived cloud validation dataset for Sentinel-2 and Landsat 8 satellites over NASA GSFC

Availability of a reliable cloud mask for optical satellite imagery is a prerequisite, when generating high-quality high-level geoinformation products. Creation of a reference (ground truth) cloud mask for moderate spatial resolution sensors, such as Operational Land Imager (OLI) aboard Landsat 8 and Multispectral Instrument (MSI) aboard Sentinel-2A/B satellites, is a challenging and time-consuming task. Existing reference datasets were mainly produced through photointerpretation of satellite images by an analyst, which can introduce subjectivity in detecting clouds. Therefore, other methods for generating cloud reference data shall be explored and evaluated that can complement existing datasets. In this paper, we document generation and provide the description of a new reference cloud dataset, named GSFC-Cloud, which is based on the extensive use of ground-based images of the sky. The dataset is collected over the same area, covers various cloud conditions, and is available for six Landsat 8 and twenty-eight Sentinel-2 scenes spanning the period of September 2017 to November 2018. The dataset is available in the vector format, so cloud masks at various spatial resolutions can be validated. We also describe a system to automate the process of ground-based data collection using low-cost off-the-shelf parts with the long-term objective to replicate this set-up in multiple locations around the world. We use the proposed dataset to validate and improve the Land Surface Reflectance Code (LaSRC) for cloud detection in Sentinel-2 imagery. We show that adding a parallax feature to estimate a subpixel shift between red and green bands with a phase correlation method can reduce overdetection of clouds and improve performance of LaSRC.

Challenging a Global Land Surface Model in a Local Socio-Environmental System

Land surface models (LSMs) predict how terrestrial fluxes of carbon, water, and energy change with abiotic drivers to inform the other components of Earth system models. Here, we focus on a single human-dominated watershed in southwestern Michigan, USA. We compare multiple processes in a commonly used LSM, the Community Land Model (CLM), to observational data at the single grid cell scale. For model inputs, we show correlations (Pearson’s R) ranging from 0.46 to 0.81 for annual temperature and precipitation, but a substantial mismatch between land cover distributions and their changes over time, with CLM correctly representing total agricultural area, but assuming large areas of natural grasslands where forests grow in reality. For CLM processes (outputs), seasonal changes in leaf area index (LAI; phenology) do not track satellite estimates well, and peak LAI in CLM is nearly double the satellite record (5.1 versus 2.8). Estimates of greenness and productivity, however, are more similar between CLM and observations. Summer soil moisture tracks in timing but not magnitude. Land surface reflectance (albedo) shows significant positive correlations in the winter, but not in the summer. Looking forward, key areas for model improvement include land cover distribution estimates, phenology algorithms, summertime radiative transfer modelling, and plant stress responses.

A Two-Step Method to Calibrate CYGNSS-Derived Land Surface Reflectivity for Accurate Soil Moisture Estimations

Based on a statistical analysis of the currently available 3-year on-orbit Cyclone Global Navigation Satellite System (CYGNSS) data (2017&#x2013;2019), this study proposes a two-step calibration method to improve the accuracy of the CYGNSS-derived land surface reflectivity (SR) and the resulting soil moisture (SM) estimates. The method is designed for two purposes: the one is to correct the system errors of the SR estimates induced by the calibration of the CYGNSS Version 2.1 L1B data, and the other is to eliminate vegetation attenuation in the SR of the soil layer. Mean SR corrections of &#x007E; &#x2212;0.9 and 2.2 dB are achieved through the first and second steps of calibration, respectively. This resulted in better SM estimates compared with the Soil Moisture Active Passive (SMAP) product and the <i>in situ</i> measurements, i.e., improved correlations between the CYGNSS SR and the SMAP SM (from <inline-formula> <tex-math notation="LaTeX">$R = 0.46$ </tex-math></inline-formula> to <inline-formula> <tex-math notation="LaTeX">$R = 0.74$ </tex-math></inline-formula>), improved correlations between the CYGNSS SR and the <i>in situ</i> SM (from <inline-formula> <tex-math notation="LaTeX">$R = 0.47$ </tex-math></inline-formula> to <inline-formula> <tex-math notation="LaTeX">$R = 0.62$ </tex-math></inline-formula>), and between the CYGNSS SM and the <i>in situ</i> SM with the best <inline-formula> <tex-math notation="LaTeX">$R = 0.64$ </tex-math></inline-formula>.

On the importance of considering land surface reflectance in earth system studies

On the importance of considering land surface reflectance in earth system studies