Spatial Downscaling of Land Surface Temperature Based on a Multi-Factor Geographically Weighted Machine Learning Model

Effect of landscape pattern changes and environmental indices on land surface temperature in a fragile ecosystem in southeastern Iran

Climate change and urbanization along with uncontrolled development in less developed countries have led to an increased ecosystems' thermal environment. Some factors such as environmental indices and landscape pattern changes can alter Land Surface Temperature (LST). Thus, the accurate evaluation of the relationship between these factors and LST is considered important for managing ecosystems, especially fragile ones under high stress. The southeast of Iran has witnessed many destructions in the environmental dimension in the past years. Moreover, this region has a low socio-economic situation, which increases the need to study in this region. In the present study, we used Landsat TM5 satellite images (1989), Landsat 8 OLI/TIRS ones (2019), and Google Earth Engine (GEE) system to prepare the maps of temporal-spatial LST changes, Land Use/Land Cover (LULC), and selected environmental indices including Normalized Difference Vegetation (NDVI), Built-up (NDBI), Water Indices (NDWI), Land Surface Moisture (LSM) and albedo. Then, the correlation levels of LST with the aforementioned indices were assessed by using Geographically Weighted Regression (GWR), as well as assessing LST variation following LULC change. In addition, the Moran index was used to analyze global and local spatial autocorrelation. The results represented an 8.67-degree increase in the mean LST during 1989-2019. Urban and built-up areas had a significant effect on increasing the temperature of the region. Additionally, water bodies and vegetation cover in the region were the most crucial parameters in LST reduction. All of the applied indices were strongly related to LST (>0.70), while some exhibited more correlation in each year. Further, the highest correlation of LST was observed with LSM and NDBI in 1989, as well as with NDVI and NDWI during 2019. In addition, the Moran index value reduced from 1989 to 2019 (from 0.93 to 0.89). Finally, the region rehabilitation based on sustainable development principles played an important role in the direct and indirect decrease in LST.

Urban Heat Island Intensity Changes in Guangdong-Hong Kong-Macao Greater Bay Area of China Revealed by Downscaling MODIS LST with Deep Learning

The urban heat island (UHI) effect caused by urbanization negatively impacts the ecological environment and human health. It is crucial for urban planning and social development to monitor the urban heat island effect and study its mechanism. Due to spatial and temporal resolution limitations, existing land surface temperature (LST) data obtained from remote sensing data is challenging to meet the long-term fine-scale surface temperature mapping requirement. Given the above situation, this paper introduced the ResNet-based surface temperature downscaling method to make up for the data deficiency and applied it to the study of thermal environment change in the Guangdong-Hong Kong-Macao Greater Bay Area (GBA) from 2000 to 2020. The results showed (1) the ResNet-based surface temperature downscaling method achieves high accuracy (R² above 0.85) and is suitable for generating 30 m-resolution surface temperature data from 1 km data; (2) the area of severe heat islands in the GBA continued to increase, increasing by 7.13 times within 20 years; and (3) except for Hong Kong and Macau, the heat island intensity of most cities showed an apparent upward trend, especially the cities with rapid urban expansion such as Guangzhou, Zhongshan, and Foshan. In general, the evolution of the heat island in the GBA diverges from the central urban area to the surrounding areas, with a phenomenon of local aggregation and the area of the intense heat island in the Guangzhou-Foshan metropolitan area is the largest. This study can enrich the downscaling research methods of surface temperature products in complex areas with surface heterogeneity and provide a reference for urban spatial planning in the GBA.

Improving LST Downscaling Quality on Regional and Field-Scale by Parameterizing the DisTrad Method

Many scientists have been investigating Land Surface Temperature (LST) because of its relevance in water management science due to its direct influence on the hydrological water cycle. This effect stems from being one of the most significant variables influencing evapotranspiration. One of the most important reasons for the evapotranspiration retrieved from MODIS data’s limited suitability for scheduling and planning irrigation schemes is the lack of spatial resolution. As a result, high-resolution LST is required for estimating evapotranspiration. The goal of this study is to improve the resolution of the available LST data, to improve evapotranspiration (ETa) estimation using statistical downscaling with Normalized Difference Vegetation Index (NDVI) as a predictor. The DisTrad (Disaggregation of Radiometric Surface Temperature) method was used for the LST downscaling procedure, which is based on aggregating the NDVI map to the LST map resolution and then calculating the coefficient of variation of the native NDVI map within the aggregated pixel and classifying the aggregated map into three classes: NDVI < 0.2 for the bare soil, 0.2 ≤ NDVI ≤ 0.5 for the partial vegetation, and NDVI > 0.5 for the full vegetation. DisTrad uses 25% of the pixels with the lowest coefficient of variation from each class to calculate the regression coefficients. In this work, adjustments to the DisTrad method were implemented to enhance downscaling LST and to examine the impacts of that alteration on the evapotranspiration estimation. The linear regression model was tested as an alternative to the original second-order polynomial. In using 10% of the pixels instead of the originally proposed 25% with the lowest coefficient of variation values, it is assumed that a group of pixels with a lower coefficient of variation represents a more homogeneous area, thus it gives more accurate values. The downscaled LST map retrieval was validated using Landsat 8 thermal maps (100 m). Applying the modified DisTrad approach to disaggregate Landsat LST to 30 m (NDVI resolution) yielded an R2 of 0.72 for the 10%, 0.74 for the 25% and 0.61 for the second-order polynomial lowest coefficient of variation compared to native LST Landsat, which means that 10% can be used as an alternative. Applying the downscaled LST map to estimate ETa yielded R2 0.84 in both cases, compared to ETa yielded from the native Landsat LST. These results prove that using the robust linear regression provided better results than using polynomial regression. With the downscaled Land Surface Temperature data, it was possible to create detailed ETa maps of the small agricultural fields in the test area.

Accounting for Almond Crop Water Use under Different Irrigation Regimes with a Two-Source Energy Balance Model and Copernicus-Based Inputs

Accounting for water use in agricultural fields is of vital importance for the future prospects for enhancing water use efficiency. Remote sensing techniques, based on modelling surface energy fluxes, such as the two-source energy balance (TSEB), were used to estimate actual evapotranspiration (ETa) on the basis of shortwave and thermal data. The lack of high temporal and spatial resolution of satellite thermal infrared (TIR) missions has led to new approaches to obtain higher spatial resolution images with a high revisit time. These new approaches take advantage of the high spatial resolution of Sentinel-2 (10–20 m), and the high revisit time of Sentinel-3 (daily). The use of the TSEB model with sharpened temperature (TSEBS2+S3) has recently been applied and validated in several study sites. However, none of these studies has applied it in heterogeneous row crops under different water status conditions within the same orchard. This study assessed the TSEBS2+S3 modelling approach to account for almond crop water use under four different irrigation regimes and over four consecutive growing seasons (2017–2020). The energy fluxes were validated with an eddy covariance system and also compared with a soil water balance model. The former reported errors of 90 W/m2 and 87 W/m2 for the sensible (H) and latent heat flux (LE), respectively. The comparison of ETa with the soil water balance model showed a root-mean-square deviation (RMSD) ranging from 0.6 to 2.5 mm/day. Differences in cumulative ETa between the irrigation treatments were estimated, with maximum differences obtained in 2019 of 20% to 13% less in the most water-limited treatment compared to the most well-watered one. Therefore, this study demonstrates the feasibility of using the TSEBS2+S3 for monitoring ETa in almond trees under different water regimes.

Reducing Scaling Effect on Downscaled Land Surface Temperature Maps in Heterogenous Urban Environments

The literature review indicates that a scaling effect does exist in downscaling land surface temperature (DLST) processes, and no substantial methods were specially developed for addressing it. In this research, the main aim is to develop a new method to reduce the scaling effect on DLST maps at high resolutions. A thermal component-based thermal spectral unmixing (TSU) model was modified and a multiple regression (REG) model was adopted to create DLST maps at high resolutions. A combined variance of red and NIR bands at a very high resolution with a difference image between upscaled LST and DLST was used to develop a new method. With two case data sets, LSTs at coarse resolutions were downscaled by using the modified TSU model and the REG model to create DLST results. The new method with a correction term expression (a linear model created by using a semi-empirical approach) was used to improve the DLST maps in the two case study areas. The experimental results indicate that the new method could reduce the root mean square error and the mean absolute error >30% and >33%, respectively, and thus demonstrate that the proposed method was effective and significant, especially reducing the scaling effect on DLST results at very high resolutions. The novel significance for the new method is directly reducing the scaling effect on DLST maps at high resolutions.

Generating 1 km Spatially Seamless and Temporally Continuous Air Temperature Based on Deep Learning over Yangtze River Basin, China

Air temperature is one of the most essential variables in understanding global warming as well as variations of climate, hydrology, and eco-systems. However, current products and assimilation approaches alone can provide temperature data with high resolution, high spatio-temporal continuity, and high accuracy simultaneously (refer to 3H data). To explore this kind of potential, we proposed an integrated temperature downscaling framework by fusing multiple remotely sent, model-based, and in-situ datasets, which was inspired by point-surface data fusion and deep learning. First, all of the predictor variables were processed to maintain spatial seamlessness and temporal continuity. Then, a deep belief neural network was applied to downscale temperature with a spatial resolution of 1 km. To further enhance the model performance, calibration techniques were adopted by integrating station-based data. The results of the validation over the Yangtze River Basin indicated that the average Pearson correlation coefficient, RMSE, and MAE of downscaled temperature achieved 0.983, 1.96 °C, and 1.57 °C, respectively. After calibration, the RMSE and MAE were further decreased by ~20%. In general, the results and comparative analysis confirmed the effectiveness of the framework for generating 3H temperature datasets, which would be valuable for earth science studies.

Spatial Downscaling of Land Surface Temperature over Heterogeneous Regions Using Random Forest Regression Considering Spatial Features

Land surface temperature (LST) is one of the crucial parameters in the physical processes of the Earth. Acquiring LST images with high spatial and temporal resolutions is currently difficult because of the technical restriction of satellite thermal infrared sensors. Downscaling LST from coarse to fine spatial resolution is an effective means to alleviate this problem. A spatial random forest downscaling LST method (SRFD) was proposed in this study. Abundant predictor variables—including land surface reflection data, remote sensing spectral indexes, terrain factors, and land cover type data—were considered and applied for feature selection in SRFD. Moreover, the shortcoming of only focusing on information from point-to-point in previous statistics-based downscaling methods was supplemented by adding the spatial feature of LST. SRFD was applied to three different heterogeneous regions and compared with the results from three classical or excellent methods, including thermal image sharpening algorithm, multifactor geographically weighted regression, and random forest downscaling method. Results show that SRFD outperforms other methods in vision and statistics due to the benefits from the supplement of the LST spatial feature. Specifically, compared with RFD, the second-best method, the downscaling results of SRFD are 10% to 24% lower in root-mean-square error, 5% to 20% higher in the coefficient of determination, 11% to 25% lower in mean absolute error, and 4% to 17% higher in structural similarity index measure. Hence, we conclude that SRFD will be a promising LST downscaling method.