It is important to accurately assess agricultural drought because of its harmful impacts on the ecosystem and economy. Soil moisture reanalysis datasets provide an important way to assess agricultural drought. In this study, the ERA5-Land surface and subsurface soil moisture was used to estimate the soil water deficit index (SWDI) in four southern provinces of China. The ERA5-Land dataset was evaluated with in situ soil moisture observations from agrometeorological stations. Agricultural drought was assessed for three climate zones at a weekly scale from 2017 to 2019 and was compared with the atmospheric water deficit (AWD). It was found that both ERA5-Land soil moisture and the derived SWDI have relatively high accuracy, and the wet bias in the ERA5-Land dataset can be reduced by the calculation of the SWDI. The subsurface layer has better performance than the surface layer in drought monitoring, though they are highly correlated. Different climate zones demonstrate different drought periods and drought severity, and the temperate climate zone with no dry season has less droughts. The most severe droughts with the largest spatial extent occurred in the early winter, especially in 2019. Differences in the SWDI and AWD are mainly shown in southwestern Yunnan. The results of this study have important reference values for drought risk management.

Soil moisture (SM) is a key variable in Earth system science that affects various hydrological and agricultural processes. Convolutional long short-term memory (Conv-LSTM) networks are widely used deep learning models for spatio-temporal SM prediction, but they are often regarded as black boxes that lack interpretability and transparency. This study aims to interpret Conv-LSTM for spatio-temporal SM prediction in China, using the permutation importance and smooth gradient methods for global and local interpretation, respectively. The trained Conv-LSTM model achieved a high R2 of 0.92. The global interpretation revealed that precipitation and soil properties are the most important factors affecting SM prediction. Furthermore, the local interpretation showed that the seasonality of variables was more evident in the high-latitude regions, but their effects were stronger in low-latitude regions. Overall, this study provides a novel approach to enhance the trust-building for Conv-LSTM models and to demonstrate the potential of artificial intelligence-assisted Earth system modeling and understanding element prediction in the future.

Remote sensing soil moisture (SM) has been widely used in various earth science studies and applications, but their low resolution limits their usage and downscaling of them is needed. In this study, we proposed a spatial downscaling method for SM based on random forest considering soil moisture memory and mass conservation to improve downscaling performance. The lagged SM was added as a predictor to represent soil moisture memory, in addition to the regular predictors in previous downscaling studies. The Soil Moisture Active Passive (SMAP) SM data of the Pearl River Basin were used to test our downscaling method. The results show that the downscaling model obtained good performance on the test set (R² = 0.848, ubRMSE = 0.034 m³/m³ and Bias = 0.008 m³/m³). The spatial and temporal performance of the RF downscaling model can be improved by adding lagged SM variables. Downscaled data obtained can retain the information of the original SMAP SM data well and show more spatial details, and mass conservation correction is considered to be useful to eliminate systematic bias of the downscaling model. Downscaled SM achieved acceptable performance in in situ validation, though it was inevitably limited by the performance of the original SMAP data. The proposed downscaling method can serve as a powerful tool for the development of high-resolution SM information.

Global carbon fluxes describe the carbon exchange between land and atmosphere. However, already available global carbon fluxes datasets have not been adjusted by the available site data and deep learning tools. In this work, a global carbon fluxes dataset (named as GCFD) of gross primary productivity (GPP), terrestrial ecosystem respiration (RECO), and net ecosystem exchange (NEE) has been developed via a deep learning based convolutional neural network (CNN) model. The dataset has a spatial resolution of 1 km at three time steps per month from January 1999 to June 2020. Flux measurements were used as a training target while remote sensing of vegetation conditions and meteorological data were used as predictors. The results showed that CNN could outperform other commonly used machine learning methods such as random forest (RF) and artificial neural network (ANN) by leading to satisfactory performance with R² values of the validation stage as 0.82, 0.72 and 0.62 for GPP, RECO, and NEE modelling, respectively. Thus, CNN trained using reanalysis meteorological data and remote sensing data was chosen to produce the global dataset. GCFD showed higher accuracy and more spatial details than some other global carbon flux datasets with reasonable spatial pattern and temporal variation. GCFD is also in accordance with vegetation conditions detected by remote sensing. Owing to the obtained results, GCFD can be a useful reference for various meteorological and ecological analyses and modelling, especially when high resolution carbon flux maps are required.

Accurate assessment of agricultural drought is useful for ecosystem services. This is a successive work of our previous study that assessed agricultural drought using the soil water deficit index (SWDI) based on ERA5-Land in the four southern provinces of China from 2017 to 2019. Firstly, in addition to ERA5-Land, the suitability of CLDAS (China Land Data Assimilation System) soil moisture for drought assessment was investigated. Then, the study was extended with more comprehensive analysis and a much longer period (1981–2020). Based on three climate zones, in situ soil moisture was used for evaluation of both reanalysis datasets and agricultural drought. It was found that ERA5-Land_SWDI and CLDAS_SWDI have a good correlation with the in situ SWDI. ERA5-Land and CLDAS demonstrate some differences in representing agricultural drought but have a similar performance evaluated by in situ soil moisture. Droughts from 2001 to 2010 were more serious than in the other three decades, and droughts have become longer and severer in some areas in the last 40 years. There was a good correlation between agricultural drought and meteorological drought. Our work offers important insights for agricultural drought risk management in the four southern provinces of China.

Soil moisture (SM) has significant impacts on the Earth’s energy and water cycle system. Remote sensing, such as the Soil Moisture Active Passive (SMAP) mission, has delivered valuable estimations of global surface soil moisture. However, it has a 2~3 days revisit time leading to gaps between SMAP areas. To achieve accurate and comprehensive real-time forecast of SM, we propose a spatial–temporal deep learning model based on the Convolutional Gated Recursive Units with Data Integration (DI_ConvGRU) to capture the spatial and temporal variation in SM simultaneously by modeling the influence of adjacent SM values in space and time. Experiments show that the DI_ConvGRU outperforms the ConvGRU with Linear Interpolation (interp_ConvGRU) and the Long Short-Term Memory with Data Integration (DI_LSTM). The best performance (Bias = 0.0132 m³/m³, ubRMSE = 0.022 m³/m³, R = 0.977) has been achieved through the use of spatial–temporal deep learning model and Data Integration term. In comparison with interp_ConvGRU and DI_LSTM, DI_ConvGRU has improved the model performance in 74.88% and 68.99% of the regions according to RMSE, respectively. The predictability of SM depends highly on SM memory characteristics. DI_ConvGRU can provide accurate spatial–temporal forecast for SM with missing data, making them potentially useful for applications such as filling observational gaps in satellite data.