Intercomparison and validation of ERA5, ERA5-LAND and AROME against in-situ measurements for wind energy applications over Reunion Island
Abstract
To ensure energy autonomy, initiatives aiming at mitigating emissions of greenhouse gases in the energy sector rely on the development of renewable energy sources (RES). However, integrating electric RES into the energy mix represents an important challenge due to the variations they undergo, which induce variations in electricity production that are not always in phase with demand. The key goal of the present study is to provide a wind resource assessment over Reunion, a small island located in the South West Indian Ocean (SWIO) aiming to become self-sufficient for its electricity with RES by 2030. Currently, wind energy comes from two farms. An increased reliance on wind energy could help reach the 100% RES target, including new offshore and onshore wind turbine capacity installations. In this study, the wind data from a variety of gridded datasets, including climate reanalysis ERA5 at ~25 km spatial resolution and ERA5-Land at ~9 km spatial resolution, along with Météo-France AROME model at ~2.5 km spatial resolution are compared and validated against in-situ measurements from different Météo-France sites across Reunion over the period 2017–2020. All datasets are recorded at 10 m height with an hourly temporal resolution. The validation was performed over different time scales, from annual means to diurnal cycles, and statistical metrics such as correlation, bias, and root-mean square error were computed to measure errors. Logarithmic law is used to extrapolate wind speed at hub heights in order to obtain the wind power density at those heights. The results reveal that the AROME is highly accurate and reliable, especially in complex terrain where lower biases are obtained in comparison to the reanalysis datasets. The wind power potential is high at 50 m and 100 m, especially in the northeast parts of the island, which appear to be suitable for onshore wind farm installation.