MODIS and VIIRS Pixel Size Discrepancy between FIRMS Viewer and reported scan and track

[cpollard]

Hi!
We've been integrating with the data provided by the MODIS_SP and VIIRS_* APIs. As per the documentation, we have been drawing the detections as rectangles centered on the lat lon of the detection with a width and height provided by the scan and track. This works fairly well, however viewing the same data on the FIRMS website yields a different result.

More specifically, it looks like the scan and track do not fully represent the data on the FIRMS viewer. As an example, I have attached an image that shows a detected pixel from the MODIS Aqua layer. The scan and track are listed as 1.9 and 1.4 km, however using the measure tool on the same website yields a result of .92 and .69 respectively. This discrepancy increases on other pixels that are only .37 km wide, while listed as the same 1.4 km track.

From our documentation reading, we have not found a reason for this difference yet, but I'm sure we are missing something obvious. Perhaps some kind of fusion is occuring on the FIRMS frontend that yields these smaller than 1 km pixels, even at NADIR.

Thanks in advance!

[LANCE & FIRMSx - ddavies]

Thanks for getting in touch and pointing this out. It looks like a scaling issue with latitude - we will look in to it more and work on a fix.