POF - Prognostic Ozone for ICON: Enabling UV Forecasts

POF - A prognostic ozone scheme for ICON: Enabling UV predictions. 

Motivation: 

Solar ultraviolet (UV) radiation at Earth’s surface poses a well-documented risk for human health. The World Health Organization has defined the UV-Index to quantify the amount of UV radiation as integers in a range of 1 to 10+. It depends on a variety of quantities, including the overhead ozone column. The UV-Index is typically forecast on the time scale of days to warn the public of health risks. In Germany this is done by the Deutscher Wetterdienst (DWD) who use their weather model ICON (ICOsahedral Nonhydrostatic model) in combination with external datasets for ozone forecasts and UV radiation calculations.  

Project:

This project set up a novel setup to make the UV-Index forecast more self-consistent by using ozone calculations via the LINearized OZone (LINOZ) scheme and UV radiation calculations via the Cloud-J scheme from within ICON and the coupled Aerosols and Reactive Trace gases (ART) module. 

Results: 

The project yielded a setup that can forecast ozone and UV-Index fields for a time frame of January to April 2025 with a precision of +-1 units of UV-Index for 94.9% of the data points in comparison to ground measurement stations. Ozone columns stay within 5% agreement for a time period of four months in the northern hemisphere in comparison to Ozonewatch satellite data. The solar zenith angle is found to be the quantity that introduces most variability (9.2 +- 1.9) on the UV-Index. Aerosol optical depth (4.1 +- 2.9), cloud cover (4.7 +- 2.1) and overhead ozone (4.9 +- 1.9) introduce smaller variabilities while the effect of surface albedo (3.3 +- 2.9) and altitude (2.5 +- 2.5) is even less pronounced. A comparison of the novel setup to the operational forecast by the DWD agrees within +-2 units of UV-Index for almost all data points with the exception mountainous areas, where larger differences occur.