Soil hydraulic parameters estimated from satellite information through data assimilation
Soil hydraulic parameters estimated from satellite information through data assimilation
复制标题
DOI：
10.1080/01431161.2010.532170
复制
发表时间：
2011-12
影响因子：
3.4
通讯作者：
Sujittra Charoenhirunyingyos;K. Honda;D. Kamthonkiat;A. Ines
中科院分区：
文献类型：
--
作者：
Sujittra Charoenhirunyingyos;K. Honda;D. Kamthonkiat;A. Ines
利用卫星观测的叶面积指数（LAI）和实际蒸散量（ETa），采用土壤水-气-植物-遗传算法（SWAP-GA）联合模型，对4个土层（60 cm以下）的土壤水力参数进行了同步估算。该反演模型通过寻找最合适的土壤水力参数集来同化遥感LAI和/或ETA，该参数集可以最小化观测和模拟LAI（LAIsim）或模拟ETA（ETasim）之间的差异。模拟土壤水分估计来自土壤水力参数进行了验证，使用安装在现场的土壤水分传感器获得的值。结果表明，由LAI单独计算的土壤水分参数能较好地估计3cm深度的土壤水分，由LAI和ETa联合计算的土壤水分参数能较好地估计12 cm深度的土壤水分，由ETa单独计算的土壤水分参数能较好地估计28 cm深度的土壤水分。似乎与60 cm深度处的测量值不匹配。因此，需要更多的信息来更好地估计更深处的土壤水力参数。尽管单靠卫星数据无法提供最低深度土壤湿度的可靠估计，但利用遥感方法推导土壤水力参数仍然是一个有前途的研究领域，具有重大的应用潜力。在农业用水管理和区域一级的水灾或旱灾预报方面尤其如此。
Leaf area index (LAI) and actual evapotranspiration (ETa) from satellite observations were used to estimate simultaneously the soil hydraulic parameters of four soil layers down to 60 cm depth using the combined soil water atmosphere plant and genetic algorithm (SWAP–GA) model. This inverse model assimilates the remotely sensed LAI and/or ETa by searching for the most appropriate sets of soil hydraulic parameters that could minimize the difference between the observed and simulated LAI (LAIsim) or simulated ETa (ETasim). The simulated soil moisture estimates derived from soil hydraulic parameters were validated using values obtained from soil moisture sensors installed in the field. Results showed that the soil hydraulic parameters derived from LAI alone yielded good estimations of soil moisture at 3 cm depth; LAI and ETa in combination at 12 cm depth, and ETa alone at 28 cm depth. There appeared to be no match with measurement at 60 cm depth. Additional information would therefore be needed to better estimate soil hydraulic parameters at greater depths. Despite this inability of satellite data alone to provide reliable estimates of soil moisture at the lowest depth, derivation of soil hydraulic parameters using remote sensing methods remains a promising area for research with significant application potential. This is especially the case in areas of water management for agriculture and in forecasting of floods or drought on the regional scale.