Incorporating vegetation dynamics noticeably improved performance of hydrological model under vegetation greening.

Numerous hydrological models calculate actual evaporation as a function of potential evaporation (PET) and soil moisture stress. There are some limitations for such empirical equations since they do not consider vegetation changes, and therefore cannot account for the different responses of soil evaporation and plant transpiration to changes in environmental factors and cannot be used for evaluating the impacts of vegetation changes. Here, we investigated whether incorporating a physically based evaporation scheme into a grid-based hydrological model can improve the accuracy of hydrological simulations. The original and modified hydrological models were evaluated in a basin which has experienced rapid vegetation greening. The model evaluations were performed using streamflow observations, soil moisture observations and water balance-based evaporation estimates. Results indicated that the modified model can provide better evaporation simulations than the original model during the period of vegetation greening. The streamflow and soil moisture simulations by the modified model over the same period benefitted significantly from the improvement in evaporation simulations and exhibited better consistency with in situ observations than the original model. This study underscores the importance of including vegetation change information in evaporation estimates and demonstrated that the physically based evaporation equation can be used in hydrological models to improve the hydrological simulations under vegetation greening conditions.

Assessment of climate change impacts on hydrology and water quality with a watershed modeling approach.

The assessment of hydrologic responses to climate change is required in watershed management and planning to protect water resources and environmental quality. This study is designed to evaluate and enhance watershed modeling approach in characterizing climate change impacts on water supply and ecosystem stressors. Soil and Water Assessment Tool (SWAT) was selected as a base model, and improved for the CO2 dependence of potential evapotranspiration and stream temperature prediction. The updated model was applied to quantify the impacts of projected 21st century climate change in the northern Coastal Ranges and western Sierra Nevada, which are important water source areas and aquatic habitats of California. Evapotranspiration response to CO2 concentration varied with vegetation type. For the forest-dominated watersheds in this study, only moderate (1-3%) reductions on evapotranspiration were predicted by solely elevating CO2 concentration under emission scenarios A2 and B1. Modeling results suggested increases in annual average stream temperature proportional to the projected increases in air temperature. Although no temporal trend was confirmed for annual precipitation in California, increases of precipitation and streamflow during winter months and decreases in summers were predicted. Decreased streamflow during summertime, together with the higher projected air temperature in summer than in winter, would increase stream temperature during those months and result in unfavorable conditions for cold-water species. Compared to the present-day conditions, 30-60 more days per year were predicted with average stream temperature >20°C during 2090s. Overall, the hydrologic cycle and water quality of headwater drainage basins of California, especially their seasonality, are very sensitive to projected climate change.