Quantifying groundwater recharge in arid and semi-arid areas is a very challenging task because recharge amounts are small in comparison with the resolution of the investigation methods. Additionally, it depends on soil characteristics, geology, topography, land cover characteristics and land use, besides being affected by the temporal variability of precipitation and other hydrometeorological variables under arid and semi-arid climates. Several watershed models used to quantify recharge generally provide recharge estimates as a residual term in the water-budget. Which means that the estimation accuracy of recharge is controlled by the measurement accuracy of the various other parameters in the water budget. Moreover, according to the literature, daily time steps are desirable for estimation of recharge because recharge generally becomes a larger component of the water budget at smaller time scales.
How could we minimize the uncertainties accompanying the estimation of groundwater recharge under semi-arid conditions using SWAT model at small scale (around 350 km2 area)? Should the SWAT use be coupled with any other techniques or models (like Hydrus-1D) to get best results? Are you aware of any studies that have used SWAT to estimate groundwater recharge, or the impact of climate change and/or anthropogenic activities on groundwater recharge under arid or semi-arid conditions?
Kiprotich Kiptum, thank you for your reply. Could you please elaborate a little bit?
Dear Mohamed Ouarani ,
There is no easy answer. My general rule of thumb in this case is, as long as you are interested in modelling surface runoff/surface processes, and the simulations match the observations, you are good to go.
However, if the focus is the estimation/quantification of groundwater processes (sur as recharge), then a broad way to ensure the value provided upon calibration is okay would be ensuring that other processes (surface runoff Q, actual evapotranspiration ET) are well represented. For example, SWAT can provide a quite accurate average annual value for surface runoff, but could underestimate ET, which might result in overestimating DEEP_AQ. Therefore, to spot this problem, having a way to diagnose ET over your study area (based on satellite data estimates (MODIS, MERRA2, etc) might help tremendously in this case. Moreover, having soil water content (SWC) estimates might also help diagnose this amount, therefore reduce the uncertainty in the estimation of recharge provided by the model.
However, if the core objective of the work is to describe groundwater processes, then a coupling might be considered. MODFLOW seems a quite popular solution in this case.
Best,
Roland.
Roland Yonaba thank you for this interesting contribution.
Bests,
Mohamed
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis