Land surface model representation of the mutual information context between multi-layer soil moisture and evapotranspiration

Authors: QUI, J. - Sun Yat-Sen University; Crow, Wade; DONG, J. - US Department Of Agriculture (USDA); NEARING, G.S. - University Of Alabama

Submitted to: Hydrology and Earth System Sciences
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/6/2020
Publication Date: 2/7/2020
Citation: Qui, J., Crow, W.T., Dong, J., Nearing, G. 2020. Land surface model representation of the mutual information context between multi-layer soil moisture and evapotranspiration. Hydrology and Earth System Sciences. 24:581–594. https://doi.org/10.5194/hess-24-581-2020.
DOI: https://doi.org/10.5194/hess-24-581-2020

Interpretive Summary: The development of microwave remote sensing techniques to measure soil moisture from space represents a major advance in our ability to globally monitor agricultural drought. However, a potential limitation in these techniques is that they are generally only sensitive to moisture content within the top 5 cm of the soil column and cannot characterize soil water availability within the entire crop root-zone (often assumed to include the top meter of the soil column). Using a unique methodology, this paper explicitly measures the information content of both surface-zone (top 5 cm) and root-zone (top 1 m) soil moisture series times for agricultural drought applications. Results demonstrate that, while root-zone soil moisture is indeed more valuable than surface-zone observations, a simple filtering transformation of surface-zone soil moisture can be performed which allows (transformed) surface-zone moisture observations to effectively match the information content of (much deeper) root-zone soil moisture observations. This result is then used to illustrate shortcomings in the representation of vertical coupling between multi-level soil moisture and surface evapotranspiration provided by existing land surface models. Insights from this paper will eventually be applied to improve these models which will, in turn, allow the USDA to better anticipate - and even mitigate - the impact of agricultural drought.

Technical Abstract: Soil moisture (') impacts the climate system by regulating incoming energy into outgoing evapotranspiration (ET) and sensible heat flux components. Therefore, investigating the coupling strength between ' and ET is important for the study of land surface/atmosphere interactions. Here, we use in-situ AmeriFlux observations to evaluate '/ET coupling strength estimates acquired from multiple land surface models (LSMs). For maximum robustness, coupling strength is represented using the sampled normalized mutual information (NMI) between ' estimates acquired at various vertical depths and surface flux represented by fraction of potential evapotranspiration (fPET, the ratio of ET to potential ET). Results indicate that LSMs are generally in agreement with AmeriFlux measurements in that surface soil moisture ('S) contains slightly more NMI with fPET than vertically integrated soil moisture ('V). Overall, LSMs adequately capture variations in NMI between fPET and ' estimates acquired at various vertical depths. However, one model – the Global Land Evaporation Amsterdam Model (GLEAM) – significantly overestimates the NMI between ' and ET and the relative contribution of 'S to total ET. This bias appears attributable to differences in GLEAM’s ET estimation scheme relative to the other two LSMs considered here (i.e., the Noah with Multi–parameterization option and the Catchment Land Surface Model). These results provide insight into improved LSM model structure and parameter optimization for land surface-atmosphere coupling analyses.