Here's the Sierra test results for this location it definitely appears to be an issue with calibration.

@RyanSpies-NOAA

Here are the results of Sierra test with different calibration scenarios.

Wow, very interesting. Can you take a look at the AHPS FIM? I wonder if it's inaccurate and causing our spatial calibration to be off.

The NWPS site does not load FIM for this location (see figure below). I checked the partner FIM locations, and this site is not included in their coverage.

Calibration points (purple dots) adjust the SRC and limit inundation to areas where we have calibration data.

When spatial calibration is turned off, the inundation map aligns better with the observed flooding in the area. This can be seen by comparing the inundation map without spatial calibration to the stage-based CatFIM map.

Ah, so it looks like there's a site on the Meramec River and these calibration points are backwater from that. @RyanSpies-NOAA don't we have some sort of logic that tries to prevent calibration points from being used in tributaries? I wonder if this one is breaking that because it's in a different HUC.

In the current version of calibration points, the points are cropped to HUC boundaries. However, in the new version, points are assigned to a HUC if the nwm lid is inside that HUC. In the figure below, the blue dots represent spatial points associated with the PCFM7 site in HUC 07140102 and are therefore assigned to HUC 07140102. There are no spatial calibration points (point observations) for HUC 07140104.

The left image shows the maximum high-flow magnitude inundation over the past 14 days for both HUCs, and the right image displays the corresponding S2 image for the same location.

This is great! Unfortunately the S2 image was probably after the peak flow for the Big River and that's why they don't match. However, your new version looks like it matches Major Stage-based CatFIM level better. Since this flood was just below major and the field has verified this location as accurate enough for NWPS, we can assume that it's a good way to compare extents.