03744nas a2200253 a 450000500170000000800410001703500150005810000200007324501380009326000670023130000340029849000270033250001420035950200590050152025500056065000180311065000300312870000440315870000550320270000450325771000640330281000590336685600650342520260818170847.0250519s20259999th u  ms t    000   eng d  a.b124562751 aBhandari, Ashok10aOptimizing land surface model for improved soil moisture estimation :bbridging the gap between simulation and satellite observations  aPathum Thani, Thailand :bAsian Institute of Technology,c2025  a173 leaves :bill.+e1 online1 aThesis; vno. WM-25-04  aA thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Water Engineering and Management  aThesis (M. Eng.) - Asian Institute of Technology, 2025  aSoil moisture is a crucial hydrologic state required for water resource management in  the country like Thailand, yet continuous and physically consistent estimates remain  limited. Land surface models offer solution for this, but optimization of these models  has been challenging due to model complexity and uncertainty. This study addresses  the problem through scheme-based optimization, including updating soil texture  parameter to achieve a balanced model. Full-factorial experiments governs the natural  selection process through switching of six distinct land surface parametrization  categories in Noah-MP v 3.6, each comprising 2-4 schemes totaling 576 experiments. Simulations were run-in high-performance computer (HPC) to identify the optimal  schemes that yields the highest mean spatial skill score (e.g, KGE = 0.58, correlation =  0.74). Results revealed the spatial variability in model performance under different  physics options and hence the tradeoff was considered for selection process. While the best performing scheme for soil moisture achieved the highest skill, it has led  to decreased performance in evapotranspiration (ET) and terrestrial water storage  anomaly (TWSA). Results from multi-variate ensemble optimization experiments  (SM+TWSA+ET) demonstrated a balanced improvement, particularly in ET  (correlation increased to 0.714) and TWSA (0.813), revealing impact across the physics  options inherent in single-variable optimization. This variation is reflected through the  under estimation of leaf area index (LAI) on wet basins and hence, is helpful for better  understanding for classification of physical parameterizations on basin scale.   The optimal model obtained through several experiments was useful for constructing  historical databases for long term trend and drought analysis. The trend analysis showed  the decline of precipitation on central and northeast basin on rainy and summer season  with average of 6.35mm /wet season. However, this seasonal trend did not align with  soil moisture and total water storages suggesting seasonal shifts. The probability  distribution for the drought obtained from Soil moisture anomaly index (SMAI) with  the long-term average from 1990 to 2023 is negatively skewed, with skewness of -0.47  with long tail on left side demonstrating the regular droughts especially on northeastern  basin. These indicators were used for regional agriculture decision support especially  on land suitability of economic crops like rice and spatial irrigation planning.   0aSoil moisture 0aEnvironmental engineering0 aNatthachet Tangdamrongsub,eChairperson1 aShanmugam, Mohana Sundaram,eExamination Committee1 aShrestha, Sangam,eExamination Committee2 aADB Japan Scholarship Program (ADB-JSP),eScholarship Donor2 aAsian Institute of Technology.tThesis; vno. WM-25-04403Full-Textuhttp://203.159.5.9/ait-thesis/detail.php?q=B22142