Thanks for the great job you’re doing. I have followed up on all the training workshops you posted on your website. It’s quite interesting, and I am eager to use the Invest model to achieve one of the objectives of my research. The model output aligned exactly with what I was looking at. (Soil erosion, sediment deposition, and sediment export downstream of the watershed) I don’t have problems with the input and output results of the mode. I want to know the following:
When I was running my model, USLE.tif was 39.9999 (ton/pixel). In the watershed attribute table (USLE_TOT) =.74,116.50(ton/yr). How to convert {ton/pixel} to {ton/year}. I am using a 30m x 30m DEM pixel, and my catchment area is 165 sq. km.
I read from the literature you provide in the workshop training materials that in data-scarce watersheds, sediment retention can be used to validate soil erosion data. I want to know if, during the validation process, I can directly correlate USLE_TOT estimated using actual LULC with USLE_TOT estimated under the assumption that in the future, if conservation measures are not enhanced, C = 1 (bare land) and P = 1.
In attribute table; there is no “Sediment retention” instead there is “avoid_exp” and “avoid_.eros”.
Hello @Elly, and welcome to the forum! I’m glad that the sediment model is working for you.
The watershed values are calculated simply by summing the per-pixel values in USLE.tif within the watershed polygon(s) you provide. The values in USLE.tif are tons/pixel/year.
I’m not quite sure what you mean by “validate”. Usually, when we think of validation, it’s related to comparing model results with observed values measured in the watershed we’re modeling. Once you calibrate the model with observed data, and validate that it’s performing acceptably, then you can feel more confident in applying the model setup to a future scenario. Do you have observed data that you can compare with?
What you’re describing sounds more like creating a scenario such that you’re comparing current conditions against a landscape that’s completely bare soil, which would provide a measure of the value of current vegetation for reducing erosion at the source (USLE).
“Sediment retention” was provided in a previous version of the model, but has been removed. We updated the model recently, and now provide “sediment deposition”, “avoided export” and “avoided erosion”, all of which are better measures of service than the old sediment retention. Unfortunately, a lot of our training material was created before we updated the model, so there are inconsistencies about the outputs that are discussed. We’ll hopefully update the training materials eventually, but that is time-consuming and we all have a lot of other projects to work on.
Thanks so much for the very detailed descriptions; they are so helpful for my research. I am working on the ungauged catchment, so I don’t have observed data. It’s like you captured all of my thoughts. I want to develop two scenarios like the ones you have highlighted above. “To compare current conditions against a landscape that’s completely bare soil”. The model is perfectly working for my work; let me try to come up with something good.