BUG: FIM cache not working properly for nearest 0.25 ft resolution

[shawncrawley]

I dug into it and it looks like the FIM cache mechanism is only half-baked as far as producing and caching FIM to the nearest 0.25ft.

The first time a specific flow is passed to HAND FIM, the extent polygon will be produced for the stage rounded to the nearest 0.25 ft, but the when it gets cached, it gets essentially labeled as the valid polygon for any flows within the stage window of the surrounding 1ft integers. For example, let's say feature 101 has flow 546cfs. Nothing is found for this in the cache, and so it gets passed to HAND FIM processing. It turns out that this flow corresponds to a stage of 4.75ft, so that extent polygon gets produced and cached. Because 4.75ft falls between the 4ft and 5ft integer stages, the polygon produced will get cached as the valid polygon for any flows that fall between the 4ft and 5ft flows. So let's say that the 4ft stage has a corresponding flow of 450cfs and that the 5ft stage has a corresponding flow of 600cfs. If the next max forecast flow at feature 101 is 451 cfs. Rather than producing a new polygon, which it should do for the 4.0ft stage, it registers that there is already one cached for that 4-5ft interval: the 4.75ft polygon. And thus it uses that one and skips the processing.

I'll fix this to process/cache appropriately.