Location: Southwest Watershed Research CenterTitle: Testing water fluxes and storage from two hydrology configurations within the ORCHIDEE land surface model across US semi-arid sites
|MACBEAN, N. - University Of Indiana|
|Scott, Russell - Russ|
|OTTLE, C. - Université Paris-Saclay|
|VUICHAR, N. - Université Paris-Saclay|
|KOLB, T. - Northern Arizona University|
|DORE, S. - Hydrofocus, Inc|
|LITVAK, M. - University Of New Mexico|
|DUCHARNE, A. - Sorbonne Universities, Paris|
|MOORE, D.J.P. - University Of Arizona|
Submitted to: Hydrology and Earth System Sciences
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/24/2020
Publication Date: 11/10/2020
Citation: Macbean, N., Scott, R.L., Biederman, J.A., Ottle, C., Vuichar, N., Kolb, T., Dore, S., Litvak, M., Ducharne, A., Moore, D. 2020. Testing water fluxes and storage from two hydrology configurations within the ORCHIDEE land surface model across US semi-arid sites. Hydrology and Earth System Sciences. 24:5203-5230. https://doi.org/10.5194/hess-24-5203-2020.
Interpretive Summary: Observations and climate models indicate that semiarid regions will likely experience more intense warming and droughts, increases in extreme rainfall events, and a greater contrast between wet and dry seasons in the future. To simulate the impact of climate change on semiarid ecosystem functioning, it is essential that the land surface model component of climate models accurately represent semiarid hydrology. We tested two different soil hydrology models to see if the next generation, more mechanistic model improved the movement and storage of water observed and measured at six semiarid sites in the southwestern United States. Our results strongly suggest that a more complex, process-based hydrology model improves daily to seasonal predictions of the root-zone soil moisture dynamics and evapotranspiration. Associated changes in the calculations of runoff, soil moisture infiltration, and bottom layer drainage also resulted in more accurate estimates of surface runoff. Therefore, the forthcoming climate model that employs this next generation hydrological model will result in more accurate and reliable simulations, especially for a large portion of the world classified as semiarid.
Technical Abstract: Water-limited semi-arid ecosystems may be particularly vulnerable to future climate change given they will likely experience more intense droughts with an increasing contrast between wet and dry seasons. Predicting future trends in moisture availability in these ecosystems relies on the ability of land surface models (LSMs) to reliably and accurately capture water stores and fluxes; however, LSM simulations have rarely been evaluated against semi-arid ecosystem in situ data. In this study, we aimed to bridge this gap by comparing soil moisture and evapotranspiration simulated by the ORCHIDEE LSM against observations from six semi-arid grass, shrub and forest sites in the southwestern USA. We tested two hydrological schemes of differing complexity: 1) a simple conceptual 2-layer bucket scheme with fixed hydrological parameters; and 2) a 11-layer mechanistic scheme of moisture diffusion in unsaturated soil based on Richard’s equations with hydrological parameters dependent on soil texture. Calculations related to surface runoff, bare soil evaporation, and water limitation have also been modified to be compatible with the different hydrology configurations. The more mechanistic 11-layer model – and associated hydrological model developments – results in simulations of upper layer soil moisture that better match the temporal variability of precipitation inputs (compared to the 2-layer bucket model). The improvement in upper layer soil moisture temporal variability results in more realistic synoptic and seasonal evapotranspiration (ET) simulations resulting from changes in both bare soil evaporation (E) and transpiration (T). The modified empirical water stress function, which is used to limit stomatal conductance, also results in a marked ET improvement (decrease) during the hot, dry pre-monsoon period (May-June). Soil moisture observations at three depths from each site further corroborate that the 11-layer model accurately describes moisture temporal dynamics in upper soil layers. However, across all sites the 11-layer model-observations agreement decreases with depth: at the high elevation forest sites the model does not capture summer storm infiltration down the soil profile, and at the low elevation shrub and grassland sites the model soil moisture is too dynamic when compared with observations. Despite improvements with the 11-layer model, discrepancies between observed and modelled soil moisture and ET remain. ORCHIDEE drastically underestimates winter/spring soil moisture and overestimates ET at the high elevation forest sites, potentially related to an underestimate of snowfall climatological forcing data resulting in a snowpack that melts too rapidly. The addition of a term that captures bare soil evaporation resistance to dry soil at the forest sites decreases E and increases T; therefore, the negatively biased 11-layer T/ET ratios are improved. At low elevation sites, the 11-layer model underestimates both the observed peak summertime ET (at shrub sites) and estimated T/ET ratios (across shrub and grass sites). The underestimate in modelled peak season ET and T/ET ratios at shrubland sites suggests a lack of transpiring leaf area – a hypothesis that is backed to some extent by the good correlation at shrub sites between model-data bias and leaf area index (LAI). The addition of the bare soil resistance term decreased summertime E across all sites, which resulted in higher year-round soil moisture content. The higher soil moisture increased T; therefore, T/ET ratios also increased – reducing the model-data bias. However, this additional term reduced model fit to ET observations across all seasons at all sites. Therefore, it is possible that the representation of semi-arid bare soil evaporation in LSMs may also need further modification, particularly at sparsely vegetated shrub and grass-dominated sites.