I am selecting an opensource hydrological model to perform a water sustainability assessment at a basin scale, including blue and green water sustainability evaluation and some considerations about water quality. I need an open source, well documented and user friendly open source software. After a first check, I identified two suitable options: SWAT+ and SWIM. I cannot really understand the differences among them. Any suggestion from someone who has already used these models? Thank you for your cooperation.
Dear Silvia,
I advise you to use SWIM or SWMM or PC-SWMM by the way, the latter is not open source, but they can give you a license for a year with the possibility of extending it for another year, especially when it comes to green technologies. SWAT is primarily intended for rural catchement.
Good luck.
Hi,
I suggest that do not use SWAT+ if you have to finish your project in the next months. Because it is under development and some bugs still are there.
But, SWAT2012 is the best, in my opinion, if you want to alter the model codes and perfect support from developers.
Best,
Mamad
Silvia De Angeli
https://www.card.iastate.edu/swat_articles/
here you could find many papers that are related to your research.
SWAT is a physically based and spatially semi-distributed model, operating on a daily time step. This model was designed to predict agriculture-related hydrological processes, including surface and soil water flow, nutrient and sediment transport, crop growth...
Brahim Abdelkebir
First of all, it is under development, and in some cases that I have used this model, I have faced lots of errors. Even, in some cases, the names of parameters are wrong in SWAT+Editor, etc.
In Google groups (SWAT+) you could find lots of discussions about.
https://groups.google.com/g/swatplus
https://groups.google.com/g/qswatplus
https://groups.google.com/g/swatplus-editor
Hi Silvia, being very succint I would say, they are similar but different....ahhhh? What? Let me try to explain.
The models are similar in the sense that they attempt to describe all relevant processes of the hydrological cycle (or most of them) on a continous surface (fully distributed) for similar time scales (e.g. daily). As with most hydrological models, each of this processes (e.g. snow melt, infiltration / efective precipitation, routing, etc) is described by what I would call a "sub-model. That being said, they use both similar and different sub-models.
To keep this short, I will give you two examples.
1. The snow melt component
>>>> SWIM uses a linear relationship ( SML = 4.57*TMX ) of the snowpack temperature -TMX- to describe daily snow melt (SML)
>>>> SWAT on the other hand uses uses a linear function of the difference between the snowpack maximum temperature and the threshold temperature for snow melt.
Here you see that they are similar in that they both assume a linear relationship between snowmelt ammount and some temperature parameter. However, they differ in the form of this relationship.
2. Runoff
>>>> Both SWIM and SWAT use the SCS (Curve Number) method to estimate runoff ammounts from precipitation with the same modification with respect to the original SCS method to account for slope effect on the ammount of runoff.
There are some slight differences between both implementations bot they are fairly close to each other.
If you go into detail on each of models (SWIM/SWAT), you will find different degrees of similarity/diference of the sub-models use to describe the different processes (or estimate a given parameter). As a side note, they have some common "ancesters" e.g. the model CREAMS, so it is understandable that they share some "genes".
You might have heard this from someone else, but you cannot say that one model is better than the other. You have to evaluate them against your needs (what exactly you want the model to do) and the resources available (time including learning curve, data) to you.
That being said, I have the feeling that SWAT might have a few more capabilities (in terms of processes and detail e.g. elevation bands treatment), but this also means that it will require more effort from your side (time/data). However, this does not assure you to obtain "better" results (whatever that may be) (PARSIMONY in big letters).
By the way, I do not know if you have already filtered out other models, but you might also want to look at The Distributed Hydrology Soil Vegetation Model -DHSVM - (https://www.pnnl.gov/projects/distributed-hydrology-soil-vegetation-model) and why not HEC-HMS. Well and there are certainly some others.
I hope this helps you out a bit.
If the study area is mainly urbanized and you are interested in smaller time steps, use SWMM. If you have a agriculture/natural dominant land use study area and daily time step for analysis is sufficient, use SWAT. If you decide to use SWAT, make sure to use its Green & Ampt runoff generation mechanism for the urban areas. SWMM has far better hydraulics compared to SWAT and hence it performs better for simulating hydrological responses for urban settings where the hydrographs show a flashier behavior.
I would recommend to use SWAT instead of SWAT+, it is a bit harder to use outside the US, but it's more reliable due to bugs in swat+ (as now, I hope It will change, as swat+ is has more user-friendly interface) - in example swat+ doesn't calculate properly urban CN (uses cn 98 value for permeable areas), overall results differ significantly from swat results using the same input (I didn't look for this bug, just gave up testing swat+).
