Cold Morning In Spots

Another morning illustrating the difficulties in forecasting low temperatures across a region with complex terrain. Above plot is from MesoWest and shows some of the reported 24-h lows ending at 8:00 am MST this morning - the range in the lows on plot spans 27 F. Here at the house the low fell to 28 F (second coldest morning since back in January). At the observations page on the NWS site I found an additional 6 stations with lows of 32 F or colder - coldest at the RAWS site on Mt. Graham (Columbine) with 19 F. The NWS grid box forecast for where the house is was for 45 F - pretty good for spots a km or 2 north and south of here but way off for the Rillito Wash. I suspect that the algorithms used to adjust the model forecast temperatures for the grids are not very sophisticated - but it's hard to find any technical details.