Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Portugal → The land-atmosphere coupling and climate extremes in Africa

Universidade de Lisboa (2017)

The land-atmosphere coupling and climate extremes in Africa

Careto, João António Martins

Titre : The land-atmosphere coupling and climate extremes in Africa

Auteur : Careto, João António Martins

Université de soutenance : Universidade de Lisboa

Grade : Mestrado em Ciências Geofísicas (Meteorologia) 2017

Soil moisture is one of the most important variables of the climate system as it constrains evapotranspiration, affecting the water and energy balances at the surface, mainly over transition areas between humid and dry climates. An analysis of the energy and moisture balance, heat waves and droughts for the Africa Coordinated Regional Downscaling Experiment (Africa-CORDEX) is performed in present climate and used to evaluate heat and moisture projections for the future. Two different RCM sets from Africa-CORDEX were exploited. One is driven by ERA-Interim reanalysis (1990-2008) and the other by Atmosphere-Ocean Global Circulation Models (AOGCMs) from the Coupled Model Intercomparison Project (CMIP5, 1960-2100), featuring the Intergovernmental Panel on Climate Change (IPCC) RCP4.5 and RCP8.5 scenarios, and focused in two periods (1971-2000 and 2071-2100). Multi-Model ensembles means were produced, for better assessing changes for the future, as well as the relationship between the fluxes partitioning and heat waves. Precipitation, Soil Moisture, Latent and Sensible Heat Fluxes, Mean, Maximum and Minimum Temperatures are also assessed and validated against observationally based databases (CRU, GPCC, FLUXNET and GLEAM) for seasonal and climatological time-scales. Overall, models display a good agreement with observations, except for some cases where strong biases are found over large areas of Africa. The multi-model ensemble is found to perform better than individual models. However, the few observations over Africa, limit the validation of individual variables and different metrics. The 10 days non-overlapping mean correlations between latent and sensible heat fluxes and, between latent and maximum temperatures are used to assess the seasonal coupling strength. Also, the Bowen ratio and the Evaporative Fraction are computed in order to evaluate the different climate evolutions and coupling for each model. Overall, all models can represent the strong soil moisture-temperature coupling regions, where those areas coincide with transition zones identified on both Bowen ratio and Evaporative Fraction. However, differences are found for individual RCMs meaning that each model represents its own reality. These strong coupling regions also correspond to regions of more heat wave events in present climate. Nonetheless, in present climate, soil moisture-temperature feedback not influence heat wave event duration, which seems to be controlled by the synoptic conditions. In future, more heat waves are expected in Africa due to an increase of mean surface temperature, but also due to changes in the spatial distribution of strong coupling regions. Drought is assessed for both Hindcast and CMIP5 simulations (Historic, RCP8.5 and RCP4.5) with the SPEI index. In the future, for both scenarios, all models agree with longer and more intense droughts over Africa. Additionally, a new coupling metric is introduced, for monthly time-scales, which considers the positive temperature extremes and the negative latent heat flux extremes. In areas where negative soil moisture anomalies influence the flux partitioning, leading to a surface increase of temperature, this metric correlate well with lower mean SPEI values for a determined period.


Version intégrale (22,72 Mb)

Page publiée le 23 avril 2018