From Wiki-WindPRO
Jump to: navigation, search


MERIT-DEM at Midtfjellet Wind Farm, Norway
An alternative and near-global DTM dataset has been developed by Dai Yanmazaki of the University of Tokio, Japan — in co-operation with collegues from the geophysical community. MERIT-DEM (Multi-Error-Removed Improved-Terrain DEM) is a dataset with near-global coverage — and it holds terrain elevations at 3 arc second resolution (approximately 90m at equator). The MERIT-DEM is based on a muliti-source approach, where SRTM3-DEM, AW3D-DEM, ViewFinder-DEM and ICESat data is used as data-input — and to filter, locate and correct the vertical errors from these spaceborne DEM's. The dataset indentifies and corrects errors from speckle noise, stripe noise, absolute bias and tree height bias. After the error removal, the percentage of global land areas mapped within 2m or better acurracy increased from 39% to 59%.

Data Evaluations

At EMD we have included the MERIT-DEM dataset in our DEM-evaluation "High Fidelity Elevation Models - What is the Value in Microscale Modelling". This study was presented at the WindEurope Wind Resource Workshop in Brussels 2019. You can find more details on model performance in the presentation powerpoint - it is avaiable as a pdf - here.

Availability in WindPRO

While the data-format provided by MERIT-DEM (geo-tiff's) can be read by WindPRO, the MERIT-DEM data is not directly available from within WindPRO due to the license conditions.

License Conditions

Due to the license conditions applied CC-BY-NC 4.0 and ODbL 1.0, we currently do not re-distribute the MERIT-DEM dataset. However, if it meets your requirements - and you have recieved an access-password from the data-providers; then please visit the MERIT-DEM website and download the tiles you need.


  • Thank you to Dai Yanmazaki of the University of Tokio, Japan, and collegues from the geophysical community - for showing the path towards a global DEM.
  • Evaluation of this dataset in the windPRO context was co-supported through the InnoWind project ( which is co-funded by the Danish Innovation Fund

External Links