Hello everyone,
So, I am facing a kind of weird problem here.
My SWAT model is situated in a nival, Arctic catchment, and significant percent of it's recharge in summer is snowmelt/icemelt.
When I was building the model, and my only water input was precipitation, the discharge output was completely different than observed values (maybe not completely different... but definetely too small).
To "fix" this problem, I used "point source" option in ArcSWAT to "imitate/mimic" few main subglacial outflows from the glacier, which are main recharge sources in the ablation season.
After doing that operation, FLOW_OUT looked much better, but now, I have the problem with calibration in SWAT-CUP.
Almost every parameter I check with the "one-at-a-time" analysis, seems to be not sensitive.
Charts look like the one attached (this example is for CN2).
Discharge charts looks good, indeed, but still baseflow and peek flows are too low compared to observed, field data.
I tried using parameters from the article suggested by Dr Karim Abbaspour (https://www.sciencedirect.com/science/article/pii/S0022169415001985) and none of them seems to be sensitive.
Any advices are higly appreciated.
Dear Krzysztof
I agreed with Christoph about using point source in SWAT model. So, choosing point source for your case might not explainable for hydrological system in your watershed.
For sensitivy analysis it is not sensible to use OAT for SWAT calibration. Because parameters are inter-dependent in watershed. So, you need to use global sensitivity analysis.
Best
Jakir
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python