I'm working with time series of weather data (specifically, min/max daily temperatures and relative humidity) recorded from weather stations over an area of about 200,000 km2.

I would be interested in spatially distributing said data, generating surfaces of weather info. This however should be done by also taking into account the diverse terrain and elevation of the region over which the weather stations are installed.

A similar project that I found referenced this work:

Thornton et al. (1997). Generating surfaces of daily meteorological variables over large regions of complex terrain. Journal of Hydrology, 190 (3-4). https://doi.org/10.1016/S0022-1694(96)03128-9

The paper presents indeed a way to generate exactly the surfaces I need. Since however my programming skills are very limited, what I would like to know is whether there already is a software, or software library (preferably free and open source), that implements the algorithm by Thornton et al. or a similar one.

I found the very promising Python library openamundsen, which seems to do exactly what I need, among many other things, but I would like to collect different possible options in order to compare them.

Any help is greatly appreciated.


1 Answer 1


I am presently working on exactly this.

I am using climate reanalysis data, to fill in the 'empties' since my climate stations are far away from each other. Or, sometime, I have N/A's in my series that I would like to fill-in or relate.

From the Copernicus site:

Climate reanalyses combine past observations with models to generate consistent time series of multiple climate variables. Reanalyses are among the most-used datasets in the geophysical sciences. They provide a comprehensive description of the observed climate as it has evolved during recent decades, on 3D grids at sub-daily intervals.

Note: While Copernicus is European, their products are global. And there are other providers of reanalyses, but Copernicus is in my opinion one of the most convenient due to:

  • their documentation,
  • the examples they provide,
  • the relatively simple API's they developed (python) and documented.
  • their tutorials to obtain data - which a non-programmer but someone technically savvy could learn to do
  • their supercomputing abilities (say you need data for your grid, you launch the request and you go through the queue of requests, and tomorrow, you got an email with your csv file or the numeric output you requested)
  • their live support (well, not live, but someone is answering nicely in a timely manner)

There are several reanalysis products available, for example:

  • ERA5
  • ERA5-Land
  • Arctic

Each have their own time interval, spatial resolution; some exclude the oceans (ERA5-Land). I personally use ERA5-Land due to the fine spatial resolution and the time interval covered. Several variables are available depending on your needs.

Note: there is other reanalysis product out there that could better fit your needs, or other solutions.

  • 1
    $\begingroup$ Thank you very much @marsisalie, I was unaware of the existence of reanalysis products, and they suit my needs much more than generic tools with which to interpolate the data myself. I'm going to take a look into the Copernicus website that you suggested. I'm leaving the question open for a few more days, in case someone suggests a tool that may be useful for other people finding this page, but then I'm selecting your answer. $\endgroup$ Oct 22, 2021 at 10:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.