Hi @laura78 -
Really sorry not to have responded to this in a timely way, but I’ll offer some thoughts now in case it’s helpful to people in the future. For the record, I’m a data analyst familiar with our models, not a hydrologist who has broader knowledge, so hydrologists might have better suggestions.
It is a big limitation of both NDR and SDR that the science behind them is based on longer-term processes, where we feel good about the results at an annual time scale, and less confident about the results at shorter time scales. If large rain events are important to capture, honestly we recommend using a different model that incorporates event-based information. That may be all you need to know.
That said, I’ll describe a process we’re testing out with SDR (which is designed very similarly to NDR), to get at seasonal distribution of sediment. This is still not event based, and adds to the uncertainty of the results, but does allow us to see general differences in rainy versus dry seasons.
We have monthly climate data, which we sum to get annual values, and run SDR with those annual values. Then we create rasters indicating the fraction of yearly precipitation that falls each month (so /). Multiplying these fraction rasters by the usle.tif or sed_export.tif model output gives the proportion of annual sediment that is produced each month.
Now, this method is very much an estimate only, and it does not address modeling particular events, but so far it’s as close as we’ve come to using our annual hydrology models at a sub-annual time scale.