Evaluating predicted smoke plume heights against airborne lidar observations for multiple fires during FIREX-AQ

Abstract:

Vertical location of biomass burning smoke at the point of emission is related to the spatial extent of the resulting impacts. While smoke injected within the planetary boundary layer (PBL) tends to influence local air quality, plumes injected into the free troposphere (FT) may negatively affect air quality at regional or even global scales and influence visibility and climate. Plume injection parameterizations exist and have been shown to improve air quality forecasts, but this problem remains ill-constrained and requires further evaluation. Here a method is presented to analyze the performance of the Freitas plume rise model implemented in WRF-Chem on wildfires sampled during the Boise deployment of the joint NASA-NOAA field campaign, Fire Influence on Regional to Global Environments and Air Quality (FIREX-AQ). The method used involves sampling the model along the flight tracks and comparing backscatter and extinction with the same parameters observed with an airborne lidar (DIAL-HSRL). Also, a method is shown to deal with fires being shifted in space and time in the model relative to observations which results in a doubled sample size. The objective of this work is to gain an understanding of when and why the plume rise parameterization succeeds or fails at predicting the location of plume injections (PBL vs FT) for multiple fires spanning different fuels and fire-size. Preliminary results show that for the model has an accuracy of over 50% but it tends to overpredict injections generating more free-tropospheric injections than observed. Results from a logistic regression analysis will be presented in order to show how modeled injection performance is related to the following variables: fire phase, time of day, area burned, fire fuels and fire radiative power (FRP). Inclusion of additional plume injection observations will also be considered. We expect this study will provide guidance on the next steps to improve plume injection parameterizations.