Page 142 - AJWEP-v22i3
P. 142

Sow and Gaye

































                Figure 1. Distribution of heat wave intensity across Senegal from 1984 to 2020. Figure created with ArcGis
                10.8, Sow and Gaye (2025).

                30°C along the coastal strip, under the influence of the   historical period 1981 – 2010. The CanESM5, with a
                Atlantic Ocean, and can reach up to 40°C in the central   spatial resolution of 2.8° × 2.8°, enables the illustration
                and eastern regions of the country, which are subjected   of the connections between heat waves and atmospheric
                to the effects of continentality. 30                blocking.   Three scenarios were selected  (SSP1-2.6,
                                                                            41
                                                                    SSP2-4.5,  and SSP6-8.5)  for two climatic  periods
                2.2. Materials                                      (2021 – 2050 and 2051 – 2080).
                To  analyze  the  future  trend  of  heat  waves,  CMIP6   The choice of the CanESM5 model was based
                simulation data were collected.  Compared to the    on a literature  review, which highlighted  its strong
                Coupled  Model  Intercomparison  Project  Phase  5   performance  in the  study of climatic  extremes,
                (CMIP5), CMIP6 introduces several improvements in   particularly due to its ability to represent the phenomena
                physical modeling and resolution. CMIP6 is the result   responsible for heat waves. Its evolution from version 2
                of two decades of research on the comparison of climate   to version  5 has enhanced  its simulation  capabilities,
                models.  It is an update  of CMIP5 that  incorporates   and its spatial resolution of 2.8° × 2.8° is particularly
                       19
                socio-economic parameters.  Several studies have used   well-suited to countries, such as Senegal, where climatic
                                        31
                CMIP6 data across various fields. 32-34  Some authors have   conditions vary significantly across regions. The choice
                concluded  that  CMIP6 performs  better  than  CMIP5   of using a single model has been adopted by Brunner
                models. 35,36  For instance,  Mmame  and  Ngongondo    et al.,  Jeong et al.,  and Schaller et al.,  allowing us
                                                               32
                                                                                      37
                                                                                                         41
                                                                         40
                report  that  CMIP6 models  provide  substantial    to focus on the specific features and results provided
                improvements in the simulation of temperature  and   by these models, which  are generally  validated  and
                precipitation,  with biases often below 10%. In this   adjusted to meet the objectives of each study.
                study,  the  CanESM5  model  was used  as  input.  This
                model  was developed  by the Canadian  Center  for   2.3. Methods
                Climate Modelling and Analysis and is available on the   2.3.1. Evaluation of the performance of the Canadian
                World Climate Research Programme (WCRP). It is an   Earth System Model Version 5
                improved version of the Second Generation Canadian   In the context of using climate model outputs, it is often
                Earth  System  Model.   Atmospheric  blocking,  which   recommended  to  conduct  evaluations  using various
                                   37
                is often the cause of heat waves,  was observed with   methods, including the Kling-Gupta Efficiency (KGE)
                                              38
                this model by Arora et al.  and Brunner et al.  over the   and Percent Bias (pBias).  Various authors, including
                                      39
                                                        40
                Volume 22 Issue 3 (2025)                       136                           doi: 10.36922/AJWEP025150107
   137   138   139   140   141   142   143   144   145   146   147