Subject: Estimating recharge from soil moisture method Posted: 8/27/2014 Viewed: 13475 times
I would like to know how you would estimate the anual groundwater recharge (or deep percolation) from the results of the 2 buckets soil moisture method?
According to what I understand I would need sum the baseflow volume (groundwater that reaches the river) plus the volume "stored" in the 2nd bucket. This stored volume seems to be only available as a % of the "deep capacity" ("relative soil moisture 2"), so I'll have to calculate the volume myself??
Any insights about that?
Thank you very much for your help!
Subject: Re: Estimating recharge from soil moisture method Posted: 8/28/2014 Viewed: 13439 times
Or should I rest "surface runoff", "evapotranspiration" and "interflow" to the total outflows to get the total groundwater recharge? This would be equal to sum "baseflow" and "increase in soil moisture", actually...
Any help on this topic??
Subject: Re: Estimating recharge from soil moisture method Posted: 9/13/2014 Viewed: 13278 times
I would suggest to use the 2-bucket MABIA method instead of the soil moisture method and let WEAP calculate the groundwater recharge.
Subject: Re: Estimating recharge from soil moisture method Posted: 9/15/2014 Viewed: 13255 times
thank you very much for your answer!
I would like to know what would be your arguments for using MABIA instead of soil moisture method. Soil moisture method can be linked to a groundwater node and act as a 2 bucket too.
Subject: Re: Estimating recharge from soil moisture method Posted: 9/17/2014 Viewed: 13219 times
the differences between MABIA and the soil moisture method are not easily described in a few words. But I'm personally convinced that MABIA is the far better choice. I understand from your first thread that you want to link the catchment to a groundwater node using an infiltration link. In that case, the second bucket of the soil moisture method would be ignored and the soil water balance model is simplified to a one bucket-approach.
But there are more arguments in favor of MABIA.
In comparison with MABIA, the soil moisture method is based on a limited range of input variables: two soil hydrological parameters such as topsoil water holding capacity and saturated hydraulic conductivity and two plant parameters Kc and ETpot. In general, the soil moisture method can be described as a simple capacity model where the water content is expressed in "percent field capacity" and soil water fluxes are driven only by the saturated hydraulic conductivity and not by potential gradients.
MABIA is a process-based soil-water balancing model calculating on daily timesteps (although the WEAP-model can still be on monthly timesteps). This gives a more realistic simulation of the soil moisture as the model can react on daily climate variation. All calculated fluxes within the soil compartments, or from the soil to groundwater are based on pedotransfer functions rather than on empirical values. Also, the growth of plants over time and thus the changes in the depth of the root zone and, as a result, evapotranspiration is automatically integrated in the simulation. In addition, MABIA supports irrigation scheduling as well as crop yield and market value analysis.
Although MABIA provides a by far bigger variety of inputs and the huge number of tabs in the data view might appear intimidating, it is (in my opinion) not more complicated to set up than a soil moisture model. WEAP-MABIA comes with a huge, editable crop library and provides excellent wizard-style tools to set up soil parameters, which make external tools, e.g. for calculating field capacities, obsolete.
Scenarios dealing with changes in landuse or irrigation techniques are very easily set up. And the results provided by MABIA are much more demonstrative than those of the soil moisture method.
Subject: Re: Estimating recharge from soil moisture method Posted: 9/17/2014 Viewed: 13212 times
Just wanted to make a few statements about the soil moisture method (SMM) relative to other models, including MABIA
which I am not too familiar with. I would argue that saying the MABIA model is "far better" is like saying chicken is better than beef. They are different models that are trying to serve similar purposes (for the most part).... In fact, I have developed a model where I blended the two approaches to take advantage of MABIA's dataset of crops; but also used the more conceptual SMM for watershed hydrology and streamflow reconstruction. We made use of soil maps and land use maps to parameterize the SMM, resulting in skillful hydrologic simulation. Also, the SMM forces the hydrologic model to "think like a the watershed"... e.g., what are its dominant process, instead of blindly following a table of parameters.
The soil moisture model was constructed in the spirit of some of Keith Bevin's articles, such as "Changing ideas in hydrology — The case of physically-based models" (1989). Models with multiple parameters give the illusion that they are better, but are these parameters real? Also, the SMM method is trying to be "accurate" but not necessarily "precise."
Just like MABIA, "the soil moisture model is a process-based soil water balance model,", but part of its novelty is that it can calculate on daily, weekly or monthly timesteps (or anything in-between). In data limited areas, this is quite useful. We modeled large basins on monthly timesteps with considerable success. We have used it just as you described, to model crop water demand but often at monthly timesteps. We use estimates of crop water demand to parameterize the soil moisture thresholds with considerable success. Indeed, we did not model "root depth"; but that is the strength of MABIA as more physically based "crop model".
Yes, the WEAP SMM is more conceptual, and like all hydrologic models (including MABIA), when used to simulate the water balance, the simulation is WRONG, but I believe can be quite USEFUL, but to be frank, doesn't pretend to know plant and crop specific parameters, which of course are gross estimates. MABIA has generally been used to model crop water demand, but not large river basin mass balances. My guess is that MABIA would be hard pressed to produce good simulations of streamflow over large basins by simply using its "pre-defined" parameter set (happy to be wrong). That said, the MABIA model is far better for modeling crop water demands on a daily time step when there is good data.
Finally, the statement that the soil water fluxes are driven only by the saturated hydraulic conductivity and not by potential gradients is not correct. By tracking relative soil moisture the vertical and horizontal fluxes are a function of the soil moisture state and an approximation of the saturated conductivity, and thus the potential gradients. Yes, the soil moisture model is arguably simple, for the reason that the range of hydrologic parameters across the landscape cannot be known, only estimated.
Happy to discuss more .. not trying be defensive.. but I have heard these arguments against the SMM for some time, and want to try and add some clarity to the discussion.
Subject: Re: Estimating recharge from soil moisture method Posted: 9/17/2014 Viewed: 13205 times
Thank you very much for your insights on both model!
I've actually been using SMM (2 buckets) to model a large watershed in Guatemala/Belize. Indeed the simple SMM feets very well the data scarcity in those country. And as David says I've managed to obtain quite accurate results (average monthly streamflow, based on average monthly precipitations) with it.
Now I just get back to my original question: how would you estimate groundwater recharge using SMM 2 buckets model? For the moment I simply estimated potential recharge resting actual evapotranspiration and surface runoff to the total precipitations. Do you suggest another way?
Subject: Re: Estimating recharge from soil moisture method Posted: 9/18/2014 Viewed: 13196 times
I didn't want to say that the one model is correct and the other is wrong. I think either model has its pros and cons and I'm sure both deliver good results if properly set up. I just wanted to describe that I personally prefer MABIA because it's more convenient to set up and has - imho - a better result presentation. But I admit that I usually deal with models in arid environments where irrigation is the most important issue. The crop and soil libraries and the irrigation schedule help a lot for my tasks.
Nevertheless, I understood that Nils wanted to retrieve separate water balances for both buckets. Isn't it correct, that the soil moisture model does not support the bottom bucket, if linked to a groundwater node? That was my main point for suggesting to use MABIA instead.
Topic "Estimating recharge from soil moisture method"