Improving the Spatial Resolution of GRACE-Derived Terrestrial Water Storage Changes in Small Areas Using the Machine Learning Spatial Downscaling Method

Gravity Recovery and Climate Experiment (GRACE) satellites can effectively monitor terrestrial water storage (TWS) changes in large-scale areas. However, due to the coarse resolution of GRACE products, there is still a large number of deficiencies that need to be considered when investigating TWS changes in small-scale areas. Hence, it is necessary to downscale the GRACE products with a coarse resolution. First, in order to solve this problem, the present study employs modeling windows of different sizes (Window Size, WS) combined with multiple machine learning algorithms to develop a new machine learning spatial downscaling method (MLSDM) in the spatial dimension. Second, The MLSDM is used to improve the spatial resolution of GRACE observations from 0.5° to 0.25°, which is applied to Guantao County. The present study has verified the downscaling accuracy of the model developed through the combination of WS3, WS5, WS7, and WS9 and jointed with Random Forest (RF), Extra Tree Regressor (ETR), Adaptive Boosting Regressor (ABR), and Gradient Boosting Regressor (GBR) algorithms. The analysis shows that the accuracy of each combined model is improved after adding the residuals to the high-resolution downscaled results. In each modeling window, the accuracy of RF is better than that of ETR, ABR, and GBR. Additionally, compared to the changes in the TWS time series that are derived by the model before and after downscaling, the results indicate that the downscaling accuracy of WS5 is slightly more superior compared to that of WS3, WS7, and WS9. Third, the spatial resolution of the GRACE data was increased from 0.5° to 0.05° by integrating the WS5 and RF algorithm. The results are as follows: (1) The TWS (GWS) changes before and after downscaling are consistent, decreasing at −20.86 mm/yr and −21.79 mm/yr (−14.53 mm/yr and −15.46 mm/yr), respectively, and the Nash–Sutcliffe efficiency coefficient (NSE) and correlation coefficient (CC) values of both are above 0.99 (0.98). (2) The CC between the 80% deep groundwater well data and the downscaled GWS changes are above 0.70. Overall, the MLSDM can not only effectively improve the spatial resolution of GRACE products but also can preserve the spatial distribution of the original signal, which can provide a reference scheme for research focusing on the downscaling of GRACE products.

Development of SOM Daily Precipitation Statistical Downscaling Model And Its Applicability Assessment in North China

A statistical downscaling method based on SOM which named SOM-SD is used over North China. It’s applicatibility by downscaling daily precipitation is evaluated. Indices are selected which represent the statistics of daily precipitation with regard to both precipitation amount (Prtot, SDII) and frequency (nr001), as well as extreme event (P95T, CWD, CDD). The large-scale predictors were extracted from the daily NCEP reanalysis data, while the predictand was high resolution gridded daily observed precipitation. A downscaling method based on SOM named SOM-SD was presented and evaluated. In evaluating, the frequency difference of wet-dry nodes is defined. And it is confirmed that there was a significant positive correlation between frequency difference and precipitation. The SOM-SD method displayed a high skill in reproducting the climatologic statistical properties of the observed precipitation. The value of BS is between 0 and 1.5×10-4. Sscore is between 0.8 and 1. The bias ranges are -7.4% and -11.6% for Prtot and SDII, -3.1days for nr001, +3.4% for P95T, -1.1 days for CWD and +3.5 days for CDD. Therefore, SOM-SD method works reasonably well.

Evaluation of the RF-Based Downscaled SMAP and SMOS Products Using Multi-Source Data over an Alpine Mountains Basin, Northwest China

Passive microwave surface soil moisture (SSM) products tend to have very low resolution, which massively limits their application and validation in regional or local-scale areas. Many climate and hydrological studies are urgently needed to evaluate the suitability of satellite SSM products, especially in alpine mountain areas where soil moisture plays a key role in terrestrial atmospheric exchanges. Aiming to overcome this limitation, a downscaling method based on random forest (RF) was proposed to disaggregate satellite SSM products. We compared the ability of the downscaled soil moisture active passive (SMAP) SSM and soil moisture and ocean salinity satellite (SMOS) SSM products to capture soil moisture information in upstream of the Heihe River Basin by using in situ measurements, the triple collocation (TC) method and temperature vegetation dryness index (TVDI). The results showed that the RF downscaling method has strong applicability in the study area, and the downscaled results of the two products after residual correction have more details, which can better represent the spatial distribution of soil moisture. The validation with the in situ SSM measurements indicates that the correlation between downscaled SMAP and in situ SSM is better than downscaled SMOS at both point and watershed scales in the Babaohe River Basin. From the TC method, the root mean square error (RMSE) of the CLDAS (CMA land data assimilation system), downscaled SMAP and downscaled SMOS were 0.0265, 0.0255 and 0.0317, respectively, indicating that the downscaled SMAP has smaller errors in the study area than others. However, the soil moisture distribution in the study area shown by the SMOS downscaled results is closer than the downscaled SMAP to the degree of drought reflected by TVDI. Overall, this study suggests that the proposed RF-based downscaling method can capture the variation of SSM well, and the downscaled SMAP products perform significantly better than the downscaled SMOS products after the accuracy verification and error analysis of the downscaled results, and it should be helpful to facilitate applications for satellite SSM products at small scales.

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.

Comparisons of Machine Learning Methods of Statistical Downscaling Method: Case Studies of Daily Climate Anomalies in Thailand

The climate change which is essential for daily life and especially agriculture has been forecasted by global climate models (GCMs) in the past few years. Statistical downscaling method (SD) has been used to improve the GCMs and enables the projection of local climate. Many pieces of research have studied climate change in case of individually seasonal temperature and precipitation for simulation; however, regional difference has not been included in the calculation. In this research, four fundamental SDs, linear regression (LR), Gaussian process (GP), support vector machine (SVM) and deep learning (DL), are studied for daily maximum temperature (TMAX), daily minimum temperature (TMIN), and precipitation (PRCP) based on the statistical relationship between the larger-scale climate predictors and predictands in Thailand. Additionally, the data sets of climate variables from over 45 weather stations overall in Thailand are used to calculate in this calculation. The statistical analysis of two performance criteria (correlation and root mean square error (RMSE)) shows that the DL provides the best performance for simulation. The TMAX and TMIN were calculated and gave a similar trend for all models. PRCP results found that in the North and South are adequate and poor performance due to high and low precipitation, respectively. We illustrate that DL is one of the suitable models for the climate change problem.

A New HASM-Based Downscaling Method for High-Resolution Precipitation Estimates

Obtaining high-quality precipitation datasets with a fine spatial resolution is of great importance for a variety of hydrological, meteorological and environmental applications. Satellite-based remote sensing can measure precipitation in large areas but suffers from inherent bias and relatively coarse resolutions. Based on the high accuracy surface modeling method (HASM), this study proposed a new downscaling method, the high accuracy surface modeling-based downscaling method (HASMD), to derive high-quality monthly precipitation estimates at a spatial resolution of 0.01° by downscaling the Integrated Multi-satellitE Retrievals for Global Precipitation Measurement (IMERG) precipitation estimates in China. A scale transformation equation was introduced in HASMD, and the initial value was set by including the explanatory variables related to precipitation. The performance of HASMD was evaluated by comparing the results yielded by HASM and the combined method of HASM, Kriging, IDW and the geographical weighted regression (GWR) method (GWR-HASM, GWR-Kriging, GWR-IDW). Analysis results indicated that HASMD performed better than the other four methods. High agreement was achieved for HASMD, with bias values ranging from 0.07 to 0.29, root mean square error (RMSE) values ranging from 9.53 mm to 47.03 mm, and R2 values ranging from 0.75 to 0.96. Compared with the original IMERG precipitation products, the downscaling accuracy with HASMD improved up to 47%, 47%, and 14% according to bias, RMSE and R2, respectively. HASMD was able to capture the spatial variation in monthly precipitation in a vast region, and it might be potentially applicable for enhancing the spatial resolution and accuracy of remotely sensed precipitation data and facilitating their application at large scales.