Under which conditions can assimilation of groundwater level data improve root zone soil moisture characterization?

The integrated model TerrSysMP couples a land surface model and a groundwater flow model, which can be used to improve the characterization of root zone soil moisture by assimilating groundwater level data. The relation between groundwater level and root zone soil moisture is affected by different climate conditions, plant functional types (PFTs) and soil textures. We assimilated groundwater level (GWLs) and surface soil moisture content (SSM) data with the Ensemble Kalman Filter (EnKF) and investigated the influence of climate conditions, vegetation types and soil properties on the assimilation performance.
The model TerrSysMP was set up for 100 synthetic cases which cover 5 climate conditions (humid temperate/humid tropical/arid tropical/ semiarid Mediterranean/ tropical monsoon), 5 plant functional types (PFTs) (needleleaf/ broadleaf/ grassland/ cropland/ bare soil) and 4 soil textures (clay/loam/ silt loam/ loamy sand), to generate different GWLs and soil moisture distributions. Data assimilation experiments are done with the 100 synthetic cases to investigate how climate conditions, PFTs and soil textures affect the joint assimilation of SSM and GWLs in terms of improving root zone soil moisture characterization.
We analyze the results of all cases by calculating the RMSE Reduction:

Under which conditions can assimilation of groundwater level data improve root zone soil moisture characterization?

We find that there is a critical zone (GWL between -5m to -1m) (see fig. 1) where assimilation results in the largest improvements in root zone soil moisture characterization, due to higher sensitivity of root zone soil moisture to the GWL in this zone. From fig.2 we see that climate conditions can greatly affect the performance of assimilation as they strongly control GWL. Soil texture also affects the assimilation performance by controlling GWL. Different soil textures have different soil hydraulic parameters which have different impacts on the joint assimilation of SSM and GWL. The PFT is found to have a relatively smaller effect on the assimilation performance.

Under which conditions can assimilation of groundwater level data improve root zone soil moisture characterization?
Fig. 1 RMSE reductions as a function of yearly averaged GWLs. Subfigures correspond to the upper 10 soil layers, each red dot represents a case, x-axis is yearly averaged GWL, y-axis is the RMSE reduction. If the RMSE reduction value is less than 1.0 (indicated by a black line), assimilation improves soil moisture estimation.
Under which conditions can assimilation of groundwater level data improve root zone soil moisture characterization?
Fig. 2 Boxplot of RMSE reductions for all cases in terms of climate conditions (upper), PFTs (middle) and soil texture types (lower). In the boxplot the red line represents median value and the green line mean value. The value 1.0 is indicated by a black dot line.

Papers involving TerrSysMP-PDAF:
Zhang, H., W. Kurtz, S.J. Kollet, H. Vereecken, and H.J. Hendricks Franssen (20xx). Comparison of different assimilation methodologies of groundwater levels to improve predictions of root zone soil moisture with an integrated terrestrial systems model. Under review for Adv. Water Resources.

Letzte Änderung: 12.12.2024