Impact of the OMI aerosol optical depth on analysis increments through coupled meteorology–aerosol data assimilation for an Asian dust storm
In this study, we investigate the possibility of using a coupled meteorology–aerosol data assimilation (DA) to improve the forecast of a dust storm. We interfaced a coupled meteorology–chemistry model (WRF-Chem) with a hybrid variational-ensemble DA system — the Maximum Likelihood Ensemble Filter (M...
Gespeichert in:
Veröffentlicht in: | Remote sensing of environment 2017-05, Vol.193, p.38-53 |
---|---|
Hauptverfasser: | , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | In this study, we investigate the possibility of using a coupled meteorology–aerosol data assimilation (DA) to improve the forecast of a dust storm. We interfaced a coupled meteorology–chemistry model (WRF-Chem) with a hybrid variational-ensemble DA system — the Maximum Likelihood Ensemble Filter (MLEF). The system is strongly coupled, in the sense that cross-covariances between meteorological and aerosol control variables are used in DA. Experiments are conducted for an Asian dust storm case that occurred over the Korean Peninsula on 13 May 2011, by assimilating both meteorological and aerosol observations. Meteorological observations include the conventional ones and the Microwave Humidity Sounder (MHS) satellite radiances, whereas aerosol observations contain aerosol optical depth (AOD) at 500nm from the Ozone Monitoring Instrument (OMI). In the control experiment (CTNL), we assimilated only meteorological observations. In the AODDA experiment, we assimilated both AOD and meteorological observations. Control variables of DA include both meteorological and aerosol ones. Our results indicate that meteorological observations can partially contribute to the change of aerosol initial conditions and vice versa, implying a cross-component impact of observations, and hence improving the utility of observations in coupled DA. The forecast comparison indicates a consistently improved fit to AOD observations after 24 h of DA. In addition, uncertainty of the 24-hour aerosol forecast in AODDA decreases significantly in local regions — by up to ∼21.6% in terms of innovations, and up to ∼81.4% in terms of square root forecast error covariance. These results are encouraging for future applications of ensemble DA with strongly coupled meteorology–chemistry modeling system.
•We apply ensemble-based coupled meteorology–aerosol data assimilation (DA).•The assimilated observations include meteorological and OMI AOD observations.•OMI AOD observations show strongly positive impact on aerosol initial conditions.•Impact of OMI AOD observations on 24-hour forecast is verified using the GOCI data.•Improvements in both analysis and forecast are noticed in consequent DA cycles. |
---|---|
ISSN: | 0034-4257 1879-0704 |
DOI: | 10.1016/j.rse.2017.02.013 |