Gigantor, thanks for generating and sharing these interesting plots.
When I see the spread in elevations for summer (0 to 5,500 ft) and winter (0 to 3,500 ft), it is hard to draw conclusions unless we know that the land area where those reports are coming from is consistent.
I know they are all from Washington State, but the range in elevations in WA State is huge per county (see attached map) and the BF reports come from a broad range of counties.
For example, BFRO reports from Kitsap County (max elevation of 1,761 ft) and Thurston County (max elevation of 2,922 ft) are not going to show that kind of broad distribution in elevation between summer and winter. Likewise, there are some counties in Eastern WA that have a narrow elevation spread (from min to max). For example, the lowest elevation in Spokane county is 1,538 ft (along the Spokane River) and the highest elevation if 5,883 ft.
Attached is a map put together by County Highpointers, showing the five WA counties with the highest standard deviation in elevation over their means. Four of them go from coast to high Cascade peaks and one goes from coast to Olympic peninsula peaks.
Taking into account the different distributions of elevation among the WA counties, then it would be interesting to check if we see any difference in the average elevation from BF reports between summer and winter for the following 3 groupings:
1) For those counties that have a narrow elevation distribution (with low elevations and close to coast).
2) For those counties that have a narrow elevation distribution (with higher elevation and away from the coast).
3) For those counties that have a broad elevation distribution (going from coast to high elevation mountains).
I think group 3 (which would include the Cascade and Olympic range counties) are the dominant group in the BFRO database and would probably show what you showed in the graphs. However, you might be able to remove the noise from those other 2 groups.