Welcome to the Earthdata Forum! Here, the scientific user community and subject matter experts from NASA Distributed Active Archive Centers (DAACs), and other contributors, discuss research needs, data, and data applications.
by thulohot » Fri Aug 04, 2023 2:15 pm America/New_York
We have recently been comparing data obtained via POWER(https://power.larc.nasa.gov/) to actual temperature measurements taken from meteorological stations. For the cases we have compared, there are major discrepancies between the data from the actual measured data and the modeled data provided through POWER. Is there a reason we are seeing such discrepancies (> 10°C) on hourly data?
Here is an example taken from the month of march 2022, with comparison taken from the Canadian historical weather data, which sources its information from (in this case) the Quebec airport meteorological station. The data from POWER was taken from the Sustainable buildings user community, which sources from MERRA-2 if I'm not mistaken.
The original csv/excel file was converted to PDF in order to attach it. Original file can be provided.
The POWER data looks to be identical to the MERRA-2 hourly data with the adjustment for local time. Since the MERRA-2 data grid box size is 0.5 latitude x 0.625 degree longitude, the average elevation for the grid box is quite a bit larger than the surface site 332 m (MERRA-2) to 74 m. That may be the main difference you are seeing in temperatures. Beyond that, local conditions can be highly variable and MERRA-2 has inherent biases both of which contribute to differences.