Impact of uncertainty in planetary boundary layer depth on concentration predictions

Accurate numerical weather prediction (NWP) and atmospheric transport and dispersion (AT&D) forecasts are essential for hazard prediction and consequence assessment. While there are many sources of error in AT&D applications, uncertainty in wind direction is generally considered the largest contributor. The next most important variables are the Planetary Boundary Layer (PBL) depth and stability. Since uncertainty in PBL depth can produce a corresponding uncertainty in estimated concentration, accurately predicting the PBL depth becomes a critical factor in determining the resulting concentration forecast.

In order to examine the effect of uncertainty in PBL depth on the downwind contaminant concentration, we examine a realistic but simple convective atmospheric boundary layer during the IHOP campaign that took place May through June 2002. Using a collection of MM5 forecasts/reanalyses that vary the physics parameterizations and include observed data, we explore the relationship between uncertainty in PBL depth and the resulting uncertainty in the concentration field. These MM5 forecasts were originally developed to create the ‘best' forecast for the given day. We use the various forecasts for PBL depth to define a realistic amount of ‘fuzziness' in that variable and therefore impact downwind concentration forecasts to some degree. We seek to relate the ‘fuzziness' in the forecast PBL depth to the resulting uncertainty in the downwind concentration field.