Running Hydropower water yield model

Hi all,

I am trying to collect data required to run Hydropower water yield model. I am facing some roadblocks while getting the data sets.
I am working on Godavari basin in India region. I am unable to find the following data sets, It would be really helpful if anyone could suggest me some data sources to collect the following data.

1) Depth to root-restricting layer- I tried collecting soil depth layer, but I could not find raster maps for this.
2) Plant available water fraction
3) Reference Evapotranspiration
I have tried with links provided in InVEST documentation pdf, But for India region I could not get data.


  • MattiasMattias Member
    edited September 2018
    you can download reference evapotranspiration from here:
    Post edited by Mattias on
  • hey Matt,

    Thanks a lot.
    Can you help me understand how to use global data set and clip it for my area of study? As I have downloaded data for PPT and ET and both are in ESRI GRID datasets(*adf file). How do I use it? I tried to clip the data in QGIS, but I couldn't. 
  • swolnyswolny Member, NatCap Staff
    Hi @Amarnath -

    For soil information, check out ISRIC Soil Grids. It has depth to bedrock, that you can use as soil depth, but you will need to calculate PAWF from soil texture data (look at the Data Sources section of the User Guide for assistance.) 

    You should be able to bring a GRID file into QGIS and clip it. What happens when you try to clip? You could try exporting the GRID file to TIFF first, then clip. TIFF is a more standard format and just easier to work with.

    ~ Stacie
  • Hi @swolny,

    Thanks for your reply,

    When I load grid and try saving it as tiff, QGIS gets crashed sometimes, or I just get a black map. And I not able to perform clipping operation. I am using QGIS 2.6.1 Brighton . 

    I have a few questions related to precipitation and ETO. I have downloaded ppt from and ET from ;

    1) I have obtained both data as 12 folders(Prec_1, Prec_2, up to Prec_12 and eto_1, eto_2, up to eto_12) So I assume that these are month wise annual averages). but for my analysis, I would require to use one file containing all 12 months in my InVEST model) Can I merge those files? IF so how do I do that?

    2) If I consider taking one file at a time say Prec_1 and eto_1 and running InVEST, would it be a problem? 

  • swolnyswolny Member, NatCap Staff
    I'm not a QGIS expert, so don't know why it is crashing on you. Have you tried a newer version of QGIS? It might be a bug that they've fixed.

    1/ Yes, those are monthly averages. You will need to add them together to get an annual average raster, and one way to do that is using the Raster Calculator tool in QGIS.

    2/ The methods used by the Hydropower water yield model are for annual average data only, so we do not recommend using monthly inputs. 

    Another thing of note is that your ET layer should be based on the same precipitation data that you're using  as input to the model. So if you want to use WorldClim precipitation, your ET layer should also be based on WorldClim precipitation data. I'm not sure which ET layer you have from Databasin, but do check how it was created so that these match.

    ~ Stacie
  • AmarnathAmarnath Member
    edited September 2018
    HI Stacie,@Swolny

    I have managed to add monthly(Precipitation and ref ET) values to get annual avg raster. 

    Regarding preparing PAWF raster, I am confused with methods to prepare it.
    One way is as u have suggested using HWSD data using REF_DEPTH field in HWSD for 'depth to root-restricting layer'.  For AWC, you can use AWC_CLASS field, turn the class into values, then divide by the soil depth to get the AWC fraction.
    Another way I discovered is by using  SoilGrids250m - Derived available soil water capacity (volumetric fraction) with FC = pF 2.0 from ;
    The latter one seems to be easy, but problem is, it is given in volumetric fraction. Can I use it directly for my PAWF file in Invest? or Do I need to divide it by soil depth? 

    Post edited by Amarnath on
  • swolnyswolny Member, NatCap Staff
    edited September 2018
    The volumetric fraction is what you need as input, you do not need to divide by soil depth. However, these are provided for 7 different depth layers, and I'm not quite sure how to handle those. @Perrine, @jandrewjohnson, do either of you have any advice?

    ~ Stacie

    Post edited by swolny on
  • PerrinePerrine Moderator, NatCap Staff
    Hi @Stacie and @Amarnath

    Sorry for the slow response here.
    The short answer is that the PAWC should be a (depth-)weighted average of PAWC values for different soil depths.
    However, note that the PAWC value will be multiplied by the soil depth or the rooting depth in the model (whichever is the smallest). Therefore, it should be representative of the soil properties over this depth.
    These values will change spatially but in general, make sure that you are not averaging over the full soil depth if the rooting depths are smaller than the soil depth (because properties beyond the rooting depth won't matter).

    Hope this makes sense!
  • Dear @Perrine,

    I would like to kindly ask how do I calculate the weighted average of PAWC values (based on depth)?

    Would it be something like this (when using the soilgrids data):
    (data_i, where i refers to the depth in cm, and 200cm is the maximum depth)

    PAWC (weighted average) = (data_0*0.00 or 0.001) + (data_5 * 5/200 ) + … 

    Or would the formula require a more complex calculation in a 3D context?

    Thank you very much!

Sign In or Register to comment.