We can't trust Cedar Breaks climate data extracted based on the site coordinates.

[teixeirak]

The UFDP is on the edge of cliff, and the valley below is 750 m lower and much warmer/drier. For any gridded data set, it is important to use a pixel that is 100% above the cliff.

Here's an explanation from Jim Lutz:

The UFDP lies in part of a PRISM pixel that also includes the very much lower area below the cliff, therefore influencing the local t/p regressions that PRISM uses (especially if you are using the 4k data set). I use the PRISM pixel directly to the north of the one that actually contains the plot (attached). This interpolation ‘issue’ would apply to any gridded data set, so just check that the pixel you are using is 100% on the rim of ’the breaks’ and does not include any of the valley (which is 750 m lower).

PRISM grid cells that I use for UFDP (800m, top and 4k, bottom). The plot is actually in the grid cell immediately south of the highlighted cell.

[teixeirak]

All data in this repo extracted from global databases are based on actual site coordinates, which means that Cedar Breaks data would be unreliable.

This issue is corrected in CRU_corrected data, but not elsewhere.

[teixeirak]

Climate summaries for UFDP are available here.

[ValentineHerr]

one easy way to deal with this would be to change the coordinates of the plot within the script, so the coordinates stay accurate in the data, but the data extracted comes from gridded layers come from the closest cell that entirely on the plateau.

I'll try to figure out what these coordinates would be for a 0.5 degree cell (what is used for CRU data).