In our lab, several remote sensing data are assimilated into models to improve simulation accuracy. The observations include leaf area index, terrestrial water storage, soil moisture, river level, and many more. With our developed assimilation technique, satellite data are assimilated both individually and simultaneously, allowing model state variables to be adjusted, leading to improved performance. The assimilation techniques enhance not only the variables directly assimilated but also the entire system, as updates are propagated throughout, resulting in the improvement of multiple variables at the same time. For instance, by assimilating satellite-derived leaf area index (LAI) into the model, the simulated LAI aligns more closely with the observations, resulting in simulation outcomes that reconcile with the satellite data (Figure 1). This may also refine evapotranspiration (ET) because LAI provides crucial information about the amount of foliage present, which directly affects the rate of ET by influencing how much water is transpired by plants and evaporated from the soil.
Figure 1: Long-term trends of leaf area index (LAI) are shown from model simulations (left), MODIS data (middle), and data assimilation estimates (right). Compared to MODIS results, the model alone fails to capture the trend variability in Thailand. Data assimilation, which integrates MODIS data into the model, produces improved LAI estimates that more closely align with the measurements.
Similarly, assimilating satellite gravity data into the model shows significant improvements in groundwater simulation, as it enhances the accuracy of water storage estimates and helps in better understanding of subsurface water dynamics. Our lab routinely seeks to advance assimilation techniques and incorporate new satellite data to maximize the impact of satellite data utilization, significantly improving our forecasts of water and climate variables.
Additionally, while univariate data assimilation (using only one type satellite observed variable) is effective for refining a specific variable, it can sometimes degrade other related variables due to the limitations in the model’s physical understanding and the potential misrepresentation of variable correlations. For example, assimilating satellite-derived soil moisture data may enhance the accuracy of the top soil layer but could inadvertently degrade the representation of other soil moisture layers or related hydrological stores. To address this challenge, our lab has developed a novel multivariate data assimilation technique that integrates a suite of satellite data into the model simultaneously. This approach ensures that multiple model variables are constrained and updated based on a broader set of observations. By incorporating diverse data sources, this technique improves the accuracy and consistency of various interconnected variables, leading to significant enhancements in overall model performance and a more comprehensive understanding of the system.
Finally, our lab aims to develop a groundbreaking AI-based data assimilation approach to overcome the computational demands and complicated implementation typically associated with physical models. This approach will enable the integration of multiple satellite observations without the extensive time and effort required to embed data assimilation schemes into traditional physical models, which often required complex recoding. Additionally, leveraging modern computational techniques, such as parallel GPU processing, will accelerate data assimilation and enhance our understanding of the system. This innovative approach promises to streamline the assimilation process, making it more efficient and scalable, while harnessing the power of advanced computational resources to achieve faster and more accurate insights.
RELATED DATA
ALICE-LAB: Asian Land Information for Climate and Environmental Research Laboratory
Asian Hydrometeorological Simulation (AHS)
Access high-resolution monthly hydrometeorological data at 1 km scale for Asia, empowering advancements in water resources management, climate research, and environmental decision-making. Unlock precise insights for optimized conservation strategies, refined hydrological forecasts, and effective climate change mitigation.
Thailand Hydrometeorological Simulation (THS)
ALICE-LAB provides daily high-resolution hydrometeorological data for Thailand at a 1 km scale, enhancing water resource management and climate research. This detailed dataset refines hydrological forecasts, supports conservation strategies, and offers critical insights for effective climate change mitigation, driving better decision-making and operational efficiency.