Water Yield Scenarios

Dear All

I am running Rios and Invest models. When I run the Rios scenarios, I saw when forest cover decrease the water vol/ha increase, than when there are transitions which could improve the soil use management. Well I would like if someone can explain me what happen?. 

I inferred that when there are less forest, there are less water evapotraspiration by plants; for that reason seems more water release, but in nature it is different.

 Thank you for you help

Annie Escobedo


  • PerrinePerrine Moderator, NatCap Staff
    Hi Annie,

    It is correct that forests consume a lot of water for evapotranspiration. As a result, the annual water yield in a forested catchment is lower than the same catchment with shorter or less dense vegetation. (This is what the InVEST annual water yield model will predict).

    The story is sometimes different at the seasonal time scale: at the end of the dry season, it is possible that a forested catchment yields more water than a grassland catchment, for example. This is because there is higher infiltration of rainwater in forested areas, which means more water to release during the dry season. This is what RIOS represents in the simple biophysical models: dry-season baseflow is increased with land use "transitions" that enhance infiltration. 

    I hope this helps,


  • Thank You Perrine.

    I have another question, How can I determine downstream beneficiaries of a watershed, it is possible? Maybe that is a political decision than spatia onel?, What do you suggest?

    Thanks a lot!
  • swolnyswolny Member, NatCap Staff
    Hi Annie -

    Downstream beneficiaries are generally based on the local condition, and if you want to combine beneficiaries with modeling results, then they'll likely have to be spatial too. Here are a couple of ways we've thought about downstream beneficiaries in the past, with regards to water yield, which might give you some ideas to work with:

    - People downstream may rely on surface water for drinking water or irrigation, and you have population information or the location of their towns and/or water intakes from the stream. With point locations for towns or intakes, you can delineate the watershed that drains into these points, and sum up the water yield within these watersheds to get the total amount of water being provided by the landscape to that point. Remember that if these watersheds overlap, the overlapping area will be providing service to multiple towns and this should be noted.  With distributed population information (density per pixel, for example), you could do a weighted Flow Length calculation, where the population is used as the weighting raster. This will essentially sum up the population downstream of each point.

    - Hydropower facilities, drinking water facilities etc may be located in the watershed, which benefit from  the water supply. Again, here we have delineated the watershed draining to the facility and summed up the water yield within the watershed. But, here it's important to note that for annual water yield, lots of forest, or reforestation, will generally lead to a reduction in water yield, which hydropower companies do not want. Usually, this comes with the tradeoff of providing more reliable baseflow during the dry season (if there is one), which they do want, but the annual model does not provide this seasonal info. There is a seasonal water yield model in the works that will better address this. So be careful if you use the annual water yield model for hydropower and other uses for which seasonal data is important. For these types of water users, it's more appropriate to quantify water production with a model that takes seasonality into account.

    ~ Stacie
Sign In or Register to comment.