An investigation of the Saharan episodes predicted by the CAMS ensemble model

Abstract

Air pollution caused by air particles is a global issue that has gotten a lot of attention because of the possible consequences for the environment and human health. Sources of particulate matter can be both natural and anthropogenic. A pollution source may only be called "natural" if it has not been polluted by human intervention. Anthropogenic particles, on the other hand, are created by human activities such as the use of fossil fuels in automobiles, power plants, home heating, and industrial operations. Arid-zone dust contributes greatly to global atmospheric aerosols. North Africa (Sahara and Sahel) is the most significant source, accounting for more than half of all worldwide dust emissions. Saharan dust mostly affects Mediterranean nations. It was calculated that in 2012–2013, it supplied over 20% (3.7 μg/m3) of the PM10 at a rural background site in Malta. The Environment and Resources Authority (ERA) of Malta is required to monitor air quality in compliance with EN 12341:2014. The Ambient Air Quality Directive (AAQD) specifies two PM10 limit levels. A daily restriction of 50 μg/m3 that cannot be exceeded more than 35 times per year, and an annual limit of 40 μg/m3. ERA is obligated by the European Commission to provide its measured data. In their Justification reports, EU Member States are expected to mention the days affected by Saharan dust. Satellite images, forecasts, and trajectories are used to do this. The CAMS Ensemble Model, however, is underutilised and understudied in the Maltese region. CAMS provides information on pollution concentrations caused by long-distance travel. It also offers free air quality dispersion forecasting for the whole European continent. As a result, if CAMS ensemble forecasting is used in Malta, it may result in more detailed forecasts. CAMS regional ensemble forecasting might potentially be used to forecast the Health Risk Index (HRI) and short-term air quality. The forecast will be used in the execution of such public accuracy since it is validated on a regular basis for the whole European region. Hence, the CAMS Ensemble Model can be a valuable tool for the ERA's verification process in identifying Saharan dust occurrences.

The primary aim of this work was to validate the CAMS ensemble model. This was achieved by doing statistical testing using the Spearman correlation test with in-situ data (ERA ground data). In which any improvements or poor model performance were remarked. Moreover, the secondary aim of this dissertation was to compare forecasting data, in-situ measurements, and ground-based observations. This was achieved by doing a descriptive analysis using the Anaconda software (python tool). In which any trends of PM10/aerosol data derived from the CAMS ensemble model (forecasting data), with MODIS satellite data (in-situ measurements), and with Environment and Resources Authority (ERA) data (ground-based observations) were noted. The primary aim showed that the CAMS Ensemble model gave generally accurate data when compared to in-situ data, with a performance accuracy of 80.3%. The only time there was no relationship between CAMS and ground data (ERA) was in 2018, when the sample size was small to analyse. As a result, this demonstrates that the CAMS Ensemble Model is a reliable tool for Saharan dust forecasting and verification procedures. The secondary aim of this study showed that ground data is the most accurate data source for comparing predicted and observed data in Saharan dust forecasts. Furthermore, due to a lack of AOD data, MODIS satellite data proved to be the least trustworthy in-situ measurement to use in this analysis. However, as demonstrated in the daily/seasonal analysis section, the MODIS combined AOD algorithm outperformed the MODIS DB AOD algorithm and some dates of the ERA`s justification report results.