The common channel attention mechanism maps feature statistics to feature weights. However, the effectiveness of this mechanism may not be assured in remotely sensing images due to statistical differences across multiple bands. This paper proposes a novel channel attention mechanism based on feature information called the feature information entropy attention mechanism (FEM). The FEM constructs a relationship between features based on feature information entropy and then maps this relationship to their importance. The Vaihingen dataset and OpenEarthMap dataset are selected for experiments. The proposed method was compared with the squeeze-and-excitation mechanism (SEM), the convolutional block attention mechanism (CBAM), and the frequency channel attention mechanism (FCA). Compared with these three channel attention mechanisms, the mIoU of the FEM in the Vaihingen dataset is improved by 0.90%, 1.10%, and 0.40%, and in the OpenEarthMap dataset, it is improved by 2.30%, 2.20%, and 2.10%, respectively. The proposed channel attention mechanism in this paper shows better performance in remote sensing land use classification.

Satellite-based soil moisture products are suitable for large-scale regional monitoring due to the accessibility. Five soil moisture products including SMAP, ESA CCI, and AMSR2 (ascending, descending, and average) were selected in the continental United States (US) from 2016 to 2021. To evaluate the performance of the products and assess their applicability, ISMN (International Soil Moisture Network) data were used as the in situ measurement. PBIAS (Percentage of BIAS), R (Pearson correlation coefficient), RMSE (Root Mean Square Error), ubRMSE (unbiased RMSE), MAE (Mean Absolute Error), and MBE (Mean Bias Error) were selected for evaluation. The performance of five products over six observation networks and various land cover types was compared, and the differences were analyzed at monthly, seasonal, and annual scales. The results show that SMAP had the smallest deviation with the ISMN data because PBIAS was around −0.13, and MBE was around −0.02 m³/m³. ESA CCI performed the best in almost all aspects; its R reached around 0.7, and RMSE was only around 0.07 m³/m³ at the three time scales. The performance of the AMSR2 products varied greatly across the time scales, and increasing errors and deviations showed from 2016 to 2020. The PBO_H2O and USCRN networks could reflect soil moisture characteristics in the continental US, while iRON performed poorly. The evaluation of the networks was closely related to spatial distributions. All products performed better over grasslands and shrublands with R, which was greater than 0.52, and ubRMSE was around 0.1 m³/m³, while products performed worse over forests, where PBIAS was less than −0.62, and RMSE was greater than 0.2 m³/m³, except for ESA CCI. From the boxplot, SMAP was close to the ISMN data with differences less than 0.004 m³/m³ between the median and lower quartiles.

Agricultural drought assessment is based on soil moisture deficit during the plant-growing season. The available long-term in situ soil moisture data can be used to evaluate the drought indices’ performance. Drought indices have different sensitivities to various climates and vegetation cover types. Moreover, little research has been carried out on the sensitivity to soil moisture in specific soil layers. Comparing the performance of drought indices in different environments enables us to use and construct an effective drought index for drought delineation. We used soil moisture data from 2000 to 2021 to compare the performance of satellite image-derived drought indices (VCI, NLSWI, TCI, VHI, and NVSWI) in different climates, vegetation cover types, and soil layers over the Contiguous United States at monthly intervals and a 1000 m spatial resolution. The performances of drought indices based on vegetation status showed little differences in the different soil layers. In contrast, drought indices based on single-temperature conditions performed better in the 0–20 cm soil layer than in the deeper soil layer. The integrated index might not improve the correlation in all soil layers compared with single-condition indices under some environments. In general, while the optimal drought index might vary in different environments, the VHI performed most efficiently in most cases. This research has the potential to improve our understanding of drought monitoring in large regions with different climates, vegetation cover types, and soil layers.