Subject: Working with a huge basin in WEAP Posted: 9/2/2009 Viewed: 40161 times
Dear all,
I would like to get the advice of anyone who has already manage a big number of WEAP catchment objects.
I am working in the Ganges basin, which has been sub-divided into 63 sub-basins. The hydrology is simulated using WEAP's catchment object. Due to the requirement of my work, some sub-basins have several catchment objects (for instance with respect to elevation, as the Himalayas are within the Ganga basin), sometime as much as 7 catchment objects per mountainous sub-basin. So far I have setted 14 sub-basins and I have 55 catchment objects. You can guess that the total number of catchment objects in the whole Ganga will be huge.
I don't have RAM problem as my system has 3 Gb plus I use a RAM Disk for the _Work folder as advised by the post http://www.weap21.org/index.asp?doc=54&action=9&read=1352&fid=30 (namely I allocate 2.2 Gb for the RAM and 800 Mb for RAM Disk). I think this works well (there seems to be no time lost in paging / SWAP).
However, WEAP starts to become very slow with my 55 catchment objects, especially when I add many scenarios (I do so for calibrating each catchment objects): WEAP seems to spend lots of time sorting all the data before starting the calculation, then even while calculating it's slow. My feeling is that WEAP is not slow to calculate neither access RAM neither write in the _Work folder, but it's slow for sorting / compiling the outputs for each catchment object / river / scenario.
To speed-up, I was thinking of calibrating each river system of the Ganga basin in a dummy WEAP area and then reverting the results into the main Ganga area, but then I guess I would spend too much time in reverting (especially inserting again the catchment object in the schematic of the Ganga area). Another option could be to create the dummy area (that I would use to calibrate a given river system of the Ganga basin) from the WEAP Ganga area but I guess that I would spend too much time deleting all the unrequired catchment objects from other river systems...
Would anyone have any magic suggestion / advice?
Thanks in advance!
Devaraj de Condappa
Mr. Jack Sieber
Subject: Re: Working with a huge basin in WEAP Posted: 9/9/2009 Viewed: 40154 times
WEAP's database tables do get slower and slower as they get larger, even on a RAMdisk, so that adding more scenarios or catchments will slow it down somewhat.
You can turn off selected catchments without needing to delete them--just set the area of a catchment to 0 and WEAP will not process it at all. What I do is just comment out the area that is entered for a catchment--put a semicolon right before the data and WEAP will treat it as a comment. If you want to get fancy, you could make a key assumption variable for each river system, which you would set to either 0 (off) or 1 (on). Then, for the area of catchments in that river system, multiply the real area by the key assumption variable. This will make it easy to turn on and off selected groups of catchments.
Best regards,
Jack
Dr. Devaraj de Condappa
Subject: Re: Working with a huge basin in WEAP Posted: 9/10/2009 Viewed: 40142 times
Thank-you very much Jack for your very useful reply.
I bounce back with a short question.
Let say that, as you proposed, I create a parameter in Key Assumption for each river system, this parameter being equal to 0 if I want to switch off calculations in this river system or 1 for the contrary. What will make WEAP calculations faster:
- to multiply the area of each catchment object in a given river system by this parameter?
- or to use an "if" condition such as "if(key assumption param = 0,0,[area of the catchment])"?
Thanks again!
Deva
Mr. Jack Sieber
Subject: Re: Working with a huge basin in WEAP Posted: 9/10/2009 Viewed: 40137 times
WEAP will not do any calculations on the catchment if the area is zero, which will be true in both cases you gave:
- to multiply the area of each catchment object in a given river system by this parameter
- or to use an "if" condition such as "if(key assumption param = 0,0,[area of the catchment])"?
Mr. Eduardo Bustos
Subject: Re: Working with a huge basin in WEAP Posted: 3/19/2021 Viewed: 10815 times
Hi Jack and Deva,
I come back to this useful post.
As a complement to these process for catchments, how you can reach the same effect, but for Groundwater element?
I'm working now in a model with also huge number of catchments, but also a substantial number of Groundwater objects. My first approach was the same. Multiply by zero the Storage Capacity parameter but the model asks for it at the beginning of the run.
The idea is the same. To calibrate in a sub region, but without need delete or deactivate each element manually.