Soil moisture (SM) is key parameter for understanding interactions between the atmosphere and Earth’s surface through energy and water cycles. The major Source of estimation of SM are land surface Models (LSM) and Satellite observations(SSM). SM estimation is hindered by spatial heterogeneity in soil, atmospheric, vegetation, and land use conditions.
There is a gap between LSM simulations and satellite observations, especially in regions like Southeast Asia like Thailand, affecting SM prediction. So, for the proper representation of land-atmosphere interactions optimization of LSM is important which will be done by adjustment of physics and parameters following satellite observation and checking the performance over other hydrological states.
Figure 1: Optimization parameters and configurations for the Noah-MP 3.6 model, incorporating SMAP satellite observations and ERA5 forcing datasets. The setup includes various physics options for vegetation models, stomatal resistance, soil moisture factors, and hydrological components, tested through 572 experimental scenarios between 2010–2023.
The study leverages the Noah-MP 3.6 model and integrates satellite data, such as SMAP, to refine the physics and parameters influencing SM predictions. Key parameters like vegetation models, stomatal resistance, and soil temperature were systematically adjusted, resulting in a marked improvement in correlation coefficients. Notably, the optimized LSM showed a 10% improvement in correlation when validated against satellite data. Seasonal analysis revealed that the monsoon season provided the best correlation, while winter had the lowest root mean square error (RMSE). Interestingly, the model performed better in Thailand’s Mun Basin but struggled in the Chao Praya Basin, highlighting the need for location-specific calibration.
Figure 2: Modeled mean soil moisture at a 10 cm depth, showcasing the spatial patterns of soil moisture (m³/m³) and their alignment with hydrological features in the study area.
Figure 3: Spatial distribution of SMAP satellite-derived mean soil moisture across the study region, illustrating varying soil moisture levels (m³/m³) during the observation period.
The research also sheds light on the limitations of the current modeling approaches. Despite advancements, areas with low soil moisture values remain challenging due to the dominance of irrigation schemes that vary significantly across regions. Furthermore, the study emphasizes the importance of tailoring physics configurations based on land use, soil characteristics, and seasonal variations for accurate hydrological predictions. These insights pave the way for more nuanced and dynamic models capable of capturing regional peculiarities.
Figure 4: Daily Mean Soil Moisture in the Chao Phraya Basin: The model predictions (solid blue line) are compared to SMAP observations (dashed red line) with a correlation of 0.6736 and an RMSE of 0.0638.
Figure 5: Daily Mean Soil Moisture in the Mun Basin: The model predictions (solid blue line) align closely with SMAP observations (dashed red line) with a correlation of 0.8963 and an RMSE of 0.0455.
The research also sheds light on the limitations of the current modeling approaches. Despite advancements, areas with low soil moisture values remain challenging due to the dominance of irrigation schemes that vary significantly across regions. Furthermore, the study emphasizes the importance of tailoring physics configurations based on land use, soil characteristics, and seasonal variations for accurate hydrological predictions. These insights pave the way for more nuanced and dynamic models capable of capturing regional peculiarities.
Further improvement on soil moisture by updating soil texture parameters and optimal soil depth for simulating surface soil moisture. Performance evaluation of improved SM parametrization over other hydrological components and respective satellite / observations is underway where the comparison is done as LAI vs GLASS,, Evapotranspiration vs GLEAM, Total water Storage vs GRACE , GPP vs MODIS, water table vs observation well, Streamflow vs Observed station. Likewise, model’s ability to capture the extremes requires to be analyzed for preparing the long-term soil moisture database and evaluate drought severity indices. This kind of database will be used for the preparation of agriculture decision support system.
Download pdf version here.
ALICE-LAB: Asian Land Information for Climate and Environmental Research Laboratory
Artificial Neural Network (ANN)
Modeled after the brain’s neural structure, ANNs are powerful tools in AI, excelling in tasks like hydrology and climate prediction by processing complex data patterns, from rainfall to extreme weather events, to enhance forecasting accuracy and resource management.
Introduction on Big Data
Big Data revolutionizes water and climate prediction by leveraging vast datasets from satellites, sensors, and historical records. Advanced analytics and machine learning improve prediction accuracy, aid in flood and drought management, and help assess climate change impacts, enabling informed decisions for a sustainable future.