Indian Forum for Water Adroit

Anybody working with VIC model

Ila Chawla

  • *****
  • Thanked: 1 times
  • +13/-0
    • View Profile
Re: Anybody working with VIC model
« Reply #60 on: February 13, 2013, 06:42:22 PM »
Hi Sat Kumar,

In input files for routing model, I am having some doubt with the UH file. In the sample files, they have given some 12 values, that is for each month they have one value and there sum should be equal to 1. Can you help me in knowing that how can that be constructed?

I have tried like from daily data I have converted into monthly, so I have 12 P's and 12 Q's then I tried using convolution method of obtaining UH ordinates and etc. But I am not getting the correct UH values. like some are negative and the first value itself is over 1.

Thanks.
 

Sat Kumar Tomer

  • *****
  • Thanked: 27 times
  • +116/-0
    • View Profile
    • ambhas
  • Institute : IISc
  • Programming language : R, Python and Matlab
Re: Anybody working with VIC model
« Reply #61 on: February 13, 2013, 09:12:26 PM »
Hi,
For the monthly scale I would have taken:
1
0
0
0
0
.....

I do not think water will stay for more than one months. If you feel you are not getting good results, you can reduce the first value a bit and increase others to make the sum equal to one.

Don't take those book method of computing UH seriously. ;)
 

Ila Chawla

  • *****
  • Thanked: 1 times
  • +13/-0
    • View Profile
Re: Anybody working with VIC model
« Reply #62 on: February 13, 2013, 09:25:05 PM »
Oh Kay...I will try this out.

Thanks for such a simple straight forward reply. I have been banging my head with Chow, raghunath and subramanya thinking that how can it be constructed... :D :D
 

Ila Chawla

  • *****
  • Thanked: 1 times
  • +13/-0
    • View Profile
Re: Anybody working with VIC model
« Reply #63 on: February 18, 2013, 10:20:00 PM »
Hi sat kumar,

Once again, I am facing a bit of problem in running routing model. When I run it, it keeps on showing searching catchment and search never gets over. But it is working very well with the test data of stehekin.

After a bit of hit and trial I found out that it is because of flow direction file. but I don't know where the problem lies. I have made the file in arc gis which gives grid values from 1 to 128. then i modified these values according to what routing model takes (As given in the sample files) and it seems to be fine. But I don't know, it doesn't run and keeps on showing searching catchment.

Did you face such a problem anytime? And can you comment what could I be missing here?

Thanks.
 

Sat Kumar Tomer

  • *****
  • Thanked: 27 times
  • +116/-0
    • View Profile
    • ambhas
  • Institute : IISc
  • Programming language : R, Python and Matlab
Re: Anybody working with VIC model
« Reply #64 on: February 20, 2013, 01:25:54 PM »
flow direction from 1 to 128?
Usually it varies from 1 to 8.
 

rajtantra

  • *
  • +1/-0
    • View Profile
  • Institute : UNBC
  • Programming language : Matlab, Python, R, C
Re: Anybody working with VIC model
« Reply #65 on: April 18, 2016, 03:25:20 AM »
Hi Sat kumar, ila,
I am facing the same problem as you have asked some time ago!

"Once again, I am facing a bit of problem in running routing model. When I run it, it keeps on showing searching catchment and search never gets over. But it is working very well with the test data of stehekin.

After a bit of hit and trial I found out that it is because of flow direction file. but I don't know where the problem lies. I have made the file in arc gis which gives grid values from 1 to 128. then i modified these values according to what routing model takes (As given in the sample files) and it seems to be fine. But I don't know, it doesn't run and keeps on showing searching catchment.

Did you face such a problem anytime? And can you comment what could I be missing here?"

Thanks.
« Last Edit: April 18, 2016, 03:28:07 AM by rajtantra »
 

rajtantra

  • *
  • +1/-0
    • View Profile
  • Institute : UNBC
  • Programming language : Matlab, Python, R, C
Re: Anybody working with VIC model
« Reply #66 on: April 18, 2016, 04:11:44 AM »
Hi Ila,
Thanks for your quick reply.
I am wondering about the flow accumulation value you mentioned here, it can be more than 128 for large watershed/area. Apart from this, I have searched a lot but did not find that people are using flow accumulation in place of flow direction file.
Do you remember how you solved the same problem?
Thanks,

Raj
 

Sat Kumar Tomer

  • *****
  • Thanked: 27 times
  • +116/-0
    • View Profile
    • ambhas
  • Institute : IISc
  • Programming language : R, Python and Matlab
Re: Anybody working with VIC model
« Reply #67 on: April 18, 2016, 02:10:08 PM »
Hi,
Please upload complete log file.

I remember routing requires flow direction map.
Kindly check basic stuff like spatial resolution, extent of study area, projection etc first.

Flow direction should be able to define a watershed. If there is some numerical error in flow direction map, it may not be able to define a watershed.
Firstly check the above stufff, if these don't help, you can upload flow direction file here. I can look for potential error in it.
 

Diwan

  • *****
  • Thanked: 32 times
  • +12/-0
  • Be a learner or Teacher or helping to learn
    • View Profile
  • Institute : National Institute of Technology Karnataka, Surathkal
  • Programming language : Python, R, Matlab
Re: Anybody working with VIC model
« Reply #68 on: April 19, 2016, 04:37:17 PM »
Dear Raj

Ya routing required flow direction and UH file. Flow direction ranges from 1 to 8...Kindly check the header of the flow direction file as mentioned by Dr. Sat... And also check the boundary grids of your watershed that flow direction should be always towards inside the flow domain. There may be a chance of looping inside the flow domain which restrict the water to flow outlet grid due to error in DEM.. Kindly check that also..   
Research Scholar
Dept. of Applied Mechanics & Hydraulics
National Institute of Technology Karnataka
Surathkal Mangalore - 575025
 

rajtantra

  • *
  • +1/-0
    • View Profile
  • Institute : UNBC
  • Programming language : Matlab, Python, R, C
Re: Anybody working with VIC model
« Reply #69 on: April 20, 2016, 01:27:36 AM »
Hi Dr. Sat and Diwan,
Thanks for your helpful suggestions. I have solved the problem. That was only because of error in flow direction file.
 

rajtantra

  • *
  • +1/-0
    • View Profile
  • Institute : UNBC
  • Programming language : Matlab, Python, R, C
Re: Anybody working with VIC model
« Reply #70 on: May 19, 2016, 03:37:12 AM »
Hi all,
I have downloaded global soil file at 0.5* resolution from the VIC website.
Is there any way to get soil file at different resolution i.e. 0.25*?

Thanks in advance!!

Raj
 

Karthikeyan L

  • *****
  • Thanked: 52 times
  • +34/-0
  • Research Scholar
    • View Profile
  • Institute : Indian Institute of Science, Bangalore
  • Programming language : Matlab, R
Re: Anybody working with VIC model
« Reply #71 on: June 16, 2016, 02:46:27 PM »
I would like to add some more information to this post. From personal communication with Sat I could learn the manner in which flow direction map should be prepared and I would like to summarize the findings here.

Once the flow direction map is prepared, though a tedious job in case of huge basins, one has to visualize how the flow is taking place from one cell to other. Say I have following 3 by 3 cells encompassing my basin.
1 1 0
1 1 0
0 0 1
where, 1 indicates locations where basin is situated and cell(3,3) is the outlet (indicating all the water in basin should drain to this cell). Let us assume we got following flow direction map.
5 7 0
5 1 0
0 0 1
where, the numbers indicate the directions provided in http://vic.readthedocs.io/en/vic.4.2.c/Documentation/Routing/FlowDirection/

If you observe cell(2,1) the flow direction, which is 5, is showing that the water is flowing outwards (south direction). And cell(2,2) has north direction. It can be observed that these numbers will form an eternal loop of water flow without it ever reaching cell(3,3). This could lead to an error while executing routing model, more specifically, "Searching for catchment" issue.
I would expect the cell(2,1)'s direction to be 3 (towards east) and cell(2,2)'s direction to be 4 (south-east towards cell (3,3)) which gives the following flow direction file.
5 7 0
3 4 0
0 0 1

Now if we visualize the flow of water from cell to cell, we can observe that water from all the cells finally reaches cell(3,3). Thank you Sat for helping me understand this very sensible concept.
 
The following users thanked this post: Ila Chawla, Diwan, rajtantra

Diwan

  • *****
  • Thanked: 32 times
  • +12/-0
  • Be a learner or Teacher or helping to learn
    • View Profile
  • Institute : National Institute of Technology Karnataka, Surathkal
  • Programming language : Python, R, Matlab
Re: Anybody working with VIC model
« Reply #72 on: June 16, 2016, 03:46:35 PM »
Dear karthikeyan
Good job.. :) nicely explained
Research Scholar
Dept. of Applied Mechanics & Hydraulics
National Institute of Technology Karnataka
Surathkal Mangalore - 575025
 
The following users thanked this post: Sat Kumar Tomer, Karthikeyan L

rajtantra

  • *
  • +1/-0
    • View Profile
  • Institute : UNBC
  • Programming language : Matlab, Python, R, C
Re: Anybody working with VIC model
« Reply #73 on: September 14, 2016, 12:51:23 AM »
Hello all,
Presently I am running the VIC model at 10Km resolution and getting the output fluxes at same resol'n.
Is there any way to interpolate the VIC outputs at fine resolution such as 5Km or 1Km??
Thank you!

Raj
 

Sat Kumar Tomer

  • *****
  • Thanked: 27 times
  • +116/-0
    • View Profile
    • ambhas
  • Institute : IISc
  • Programming language : R, Python and Matlab
Re: Anybody working with VIC model
« Reply #74 on: September 15, 2016, 03:12:09 PM »
You can as well run the VIC at fine spatial resolution.
But, it will not help much unless you improve the spatial resolution of  precipitation and soil parameters.
 
The following users thanked this post: rajtantra