Conceptualizing the Impact of Dust Contaminated Infrared Radiances on Data Assimilation for Numerical Weather Prediction

• Published in: Journal of atmospheric and oceanic technology Vol. 38; no. 2; pp. 209 - 221
• Main Authors: Marquis, Jared W., Campbell, James R, Ruston, Benjamin C., Córdoba-Jabonero, Carmen, Lewis, Jasper R, Toth, Travis D., Zhang, Jianglong
• Format: Journal Article
• Language: English
• Published: Goddard Space Flight Center American Meteorological Society 01.02.2021
• Subjects: Aerosol effects; Aerosol observations; Aerosol optical depth; Aerosol Robotic Network; Aerosols; Armed forces; Atmospheric particulates; Clouds; Contamination; Data assimilation; Data collection; Dust; Dust storms; Earth Resources And Remote Sensing; Environment Pollution; Exploitation; Infrared radiation; Innovations; Lidar; Mathematical models; Meteorological satellites; Moisture; Numerical weather forecasting; Oceanography; Optical analysis; Optical thickness; Predictions; Radiance; Radiative transfer; Rawinsondes; Sensors; Temperature; Weather forecasting; United States > US; Tenerife; Spain
• ISSN: 0739-0572; 1520-0426
• DOI: 10.1175/JTECH-D-19-0125.1

Numerical weather prediction systems depend on Hyperspectral Infrared Sounder (HIS) data, yet the impacts of dust-contaminated HIS radiances on weather forecasts has not been quantified. To determine the impact of dust aerosol on HIS radiance assimilation, we use a modified radiance assimilation system employing a one-dimensional variational assimilation system (1DVAR) developed under the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) Numerical Weather Prediction–Satellite Application Facility (NWP-SAF) project, which uses the Radiative Transfer for TOVS (RTTOV). Dust aerosol impacts on analyzed temperature and moisture fields are quantified using synthetic HIS observations from rawinsonde, Micropulse Lidar Network (MPLNET), and Aerosol Robotic Network (AERONET). Specifically, a unit dust aerosol optical depth (AOD) contamination at 550 nm can introduce larger than 2.4 and 8.6 K peak biases in analyzed temperature and dewpoint, respectively, over our test domain. We hypothesize that aerosol observations, or even possibly forecasts from aerosol predication models, may be used operationally to mitigate dust induced temperature and moisture analysis biases through forward radiative transfer modeling.