One of the largest uncertainties on air quality forecasts corresponds to smoke predictions from biomass burning emissions. The fire impacts on regional to global scales: emissions, chemistry,…

This federal contract award is a grant for $269,543.52 from the National Aeronautics and Space Administration (NASA) to the University of California, Los Angeles (UCLA) through its Office of Research Administration division. The grant supports the "FIRE impacts on regional to global scales: Emissions, Chemistry, Transport, and Models (FIRECHEM)" campaign, which aims to consolidate multiple air quality forecast systems to assess their performance and reduce uncertainties in smoke predictions from biomass burning.

The key objectives are to: 1) Evaluate the near-real-time and post-campaign performance of the forecast systems using satellite and ground-based data; 2) Provide ensemble predictions by combining the forecasts in an optimal way; and 3) Standardize the display of the forecast results to facilitate their interpretation and use in flight planning. The grant does not have a set-aside designation, and the work will be performed at UCLA's facilities in Los Angeles. The broader context is supporting NASA's efforts to improve air quality modeling and forecasting capabilities, which are critical for understanding the regional to global impacts of biomass burning.

One of the largest uncertainties on air quality forecasts corresponds to smoke predictions from biomass burning emissions. The fire impacts on regional to global scales: emissions, chemistry, transport, and models (FIRECHEM) campaign represents a unique opportunity to consolidate multiple forecast systems to assess their performance and find ways to reduce uncertainties. At the same time, flight-planning activities for FIRECHEM would greatly benefit by having all these forecasts readily available and displayed in a standardized way over the regions of interest. This would also help with the coordination with the partner field campaigns occurring simultaneously. We propose an effort to coordinate multiple forecast systems that will be available during fire-chem (10 operational/funded, 4 pending funding) to support flight planning and perform post-campaign analysis. With this effort we expect to facilitate the sampling required by the FIRECHEM objectives, help answering FIRECHEM science questions, and support coordination with partner campaigns. The following tasks will be performed: 1. Near-real-time and post-campaign evaluation of the forecasts. We plan to evaluate the predictions from the forecast systems in an automated way using satellite and ground-based data that are available in near-real-time. These performance metrics will be available at the forecasting briefings and will help determine the confidence in the forecasts when planning the flights. We will then incorporate campaign data to perform a more thorough assessment where local and international air quality forecast models, along with their respective smoke emissions, will be evaluated to help constrain north american fire emissions and assess process representation (e.g., plume injection, diurnal cycle of emissions, smoke aging) within the forecasts. The forecasts use satellite data as input at different levels (e.g., computing emissions, assimilation), thus assessing their performance will provide insights into the importance of the satellite data for north american fire prediction. 2. Ensemble predictions. Combining forecasts from multiple systems in an optimal way generally provides better predictions than each individual member and IT allows quantification of forecast uncertainty. Thus, we propose to provide an ensemble prediction using the forecast systems. We will provide statistical products such as the ensemble median, mean, spread, and probability of exceeding concentration thresholds. Also, the individual and ensemble forecasts will be improved by applying bias correction techniques based on geostationary AOD retrievals. 3. Standardization of the forecast systems. Forecasting groups during field campaigns typically plot their results in different color scales and different projections, and post them on different websites, making the interpretation of the forecasts a difficult and time consuming task. We propose to gather model outputs of key parameters from multiple forecasts and display them under the same framework to facilitate the inter-comparison in daily forecasting briefings. This will help the forecaster to focus on the analysis of the forecasts rather than on the compilation of the data. In short, during Y1 of the project we will perform a dry-run of these activities with the available forecast systems, in Y2, along with the tasks mentioned above, we will be present in the field to interpret the forecasts and participate in flight planning and, in Y3, we will summarize the forecast performance in research papers. We have assembled a team of scientists representing multiple chemical forecast systems from NASA, NOAA, ncar, NRL, eccc, ecmwf and US universities. The PI has previous collaboration with several of these groups and has participated in multiple NASA field campaigns (SEAC4RS, korus-AQ, oracles) in both forecasting and flight planning. The team also includes experts on ensemble prediction and web interfaces.