Posters

Abstract

It is very common for wind analysts to use easily online-accessible global databases for Digital Topography Models (DTM) by default, or in the absence of any local dataset. SRTM (Shuttle Radar Topographic Mission) and NASADEM (NASA Digital Elevation Model) are well-known examples of these datasets providing global topography data based on satellite measurements. These DTMs are free to use and immediately available, compared to more expensive regional datasets or expensive and time-consuming local LiDAR scans. The literature provides some information about the accuracy of these datasets. They point to potential biases of these datasets in forested and urban areas, where satellites take into account building and tree heights. However, there is little information available about the impact of this choice on energy yield and associated uncertainties. In the present work we propose to analyze this matter in the light of several case studies located in Belgium, Australia, and Peru. In the Belgium case biases of several meters were identified between heights from the NASADEM datasets and local surveys, in urban, forested, but also flooded environments. The impact on energy yield was computed, yielding biases of 3 to 4% on energy yield at some locations. The impact of local inconsistencies versus inconsistencies located further away from turbine locations was assessed. In some simple cases removing local inconsistencies manually led to a more accurate energy yield assessment. In the Australian and Peruvian cases, located in complex terrain, NASADEM and SRTM data were compared to local datasets and LiDAR scans. Positive biases were identified in forested areas primarily, leading to biases of the same order of magnitude than in the Belgium case. We conclude by discussing uncertainties associated with global databases depending on terrain and land cover type, and the added-value of in-situ LiDAR scans.