Your current browser is not supported by our site. We recommend using Chrome or Firefox to ensure full functionality.
Utah Climate Center banner

    Inversion Forecast


    Discussion Last Updated: 03-04-2016 3:07:43 pm MST

    The official end of inversion season generally arrives at the end of March. However, the exceptionally warm temperatures the past few week have allowed valley ground layers to completely thaw meaning that any remaining snowfall reaching the valley floors will be quick to melt. As the zenith angle quickly decreases, longer days lead to free convection which allows for atmospheric mixing even in the presence of strong high pressure. In short, this inversion season has come to a close with a mean troughing pattern expected to cover the region throughout March.

    This product is being developed by researchers at the Utah Climate Center. Using output from the National Center for Environmental Prediction’s (NCEP) Climate Forecast System (CFSv2), this technique projects surface inversion probability for persistent inversion events—defined as events lasting longer than 4 days—with a demonstrated “skill” over a span of ~ 30 days. A surface inversion probability of 35% or greater suggests a statistically significant likelihood of an extended event. It should also be noted that inversion forecasts, in and of themselves, are not air quality projections. The projection is valid for a radius of roughly 200 miles around Salt Lake City.

    Image Interpretation: The blue bar graph shows the calculated Surface Inversion Probabilities. Values above the horizontal yellow line (~35% on the right axis) have a statistical significance of manifesting as persistent (> 4 days) inversion events. The solid black line above the SIP chart shows the ensemble average of 200mb geopotential heights for the most recent 16 CFSv2 forecasts. The individual dotted lines are individual model runs. The vertical red and yellow lines identify the initialization (+0) and 30 day (+30) locations. SIP values lying within this 30 day window have a statistically significant confidence interval. Values before the initialization data are obtained from NCEP reanalysis data.

    For more information on the methodology and origins of this product, see the following publication here.