Tellus: Series B, Chemical and Physical Meteorology (Feb 2015)

Seasonal patterns of Saharan dust over Cape Verde – a combined approach using observations and modelling

  • Carla Gama,
  • Oxana Tchepel,
  • José María Baldasano,
  • Sara Basart,
  • Joana Ferreira,
  • Casimiro Pio,
  • João Cardoso,
  • Carlos Borrego

DOI
https://doi.org/10.3402/tellusb.v67.24410
Journal volume & issue
Vol. 67, no. 0
pp. 1 – 21

Abstract

Read online

A characterisation of the dust transported from North Africa deserts to the Cape Verde Islands, including particle size distribution, concentrations and optical properties, for a complete annual cycle (the year 2011), is presented and discussed. The present analysis includes annual simulations of the BSC-DREAM8b and the NMMB/BSC-Dust models, 1-yr of surface aerosol measurements performed within the scope of the CV-DUST Project, AERONET direct-sun observations, and back-trajectories. A seasonal intrusion of dust from North West Africa affects Cape Verde at surface levels from October till March when atmospheric concentrations in Praia are very high (PM10 observed concentrations reach hourly values up to 710 µg/m3). The air masses responsible for the highest aerosol concentrations in Cape Verde describe a path over the central Saharan desert area in Algeria, Mali and Mauritania before reaching the Atlantic Ocean. During summer, dust from North Africa is transported towards the region at higher altitudes, yielding to high aerosol optical depths. The BSC-DREAM8b and the NMMB/BSC-Dust models, which are for the first time evaluated for surface concentration and size distribution in Africa for an annual cycle, are able to reproduce the majority of the dust episodes. Results from NMMB/BSC-Dust are in better agreement with observed particulate matter concentrations and aerosol optical depth throughout the year. For this model, the comparison between observed and modelled PM10 daily averaged concentrations yielded a correlation coefficient of 0.77 and a 29.0 µg/m3 ‘bias’, while for BSC-DREAM8b the correlation coefficient was 0.63 and ‘bias’ 32.9 µg/m3. From this value, 12–14 µg/m3 is due to the sea salt contribution, which is not considered by the model. In addition, the model does not take into account biomass-burning particles, secondary pollutants and local sources (i.e., resuspension). These results roughly allow for the establishment of a yearly contribution of 42% of dust from North African deserts for PM10 levels observed in Cape Verde.

Keywords