Include NWS Near-term Forecasts with production runs

[akrherz]

Brian pointed out this news item regarding the WI DNR doing a runoff risk model based on NWS RFC data. Can something similar be done with DEP?

The primary issue is that we don't have a good source of high-temporal resolution forecast precipitation data, so getting the detachment/soil loss numbers right would be tough. But, if we only cared about runoff, having hourly or three-hourly precip input would likely be good enough. From some old sensitivity work I did, it seemed that WEPP was mostly sensitive to daily precip totals and not sub-hourly intensities when it came to runoff values.

Implementation

Currently, DEP runs for dates into the 'future' by simply repeating a previous years data to fill out the current year. So, in theory, we could have a process that merges a NWS forecast to replace the upcoming few days within the CLI files. The realtime runs ignore any dates into the future, so we could have a secondary post-processor that extracts the runoff data.

The actual computation of a 'risk' and/or visualization of this data would need further discussion.

[akrherz]

Links to nearby state sites for forecast runoff risk
MN: https://mnag.maps.arcgis.com/apps/MapSeries/index.html?appid=7a6538ffc8994715ba668f0579fe53a2
WI: http://www.manureadvisorysystem.wi.gov/runoffrisk/index