Comparing GCM data and KNN WG for having future data?

These are separate tools or methods. So you have two options:

  1. If you want to generate future data from your historical data of a station. So you can use the Artificial Intelligence method that the Name is KNN Weather Generator. By this method, you can produce future data based on the similarity of your historical data.
  2. If you want to generate data by using GCM data so you can download the files from the USGS portal(or others). These data are in NetCDF format and in coarse-scale. You have two challenges to do it. A- Netcdf format is a format of gridded data and those are all over the world. For extract data from Netcdf files, you should use programming languages such as Matlab and it's a complicit process. B- You would use a bias correction method to convert a coarse-scale data to pointy(station data).

The SD-GCM tool can simplify both challenges for using GCM data. This the link to our papers: Papers. You can find a paper in a file with a name: KNN_WG-CMIP5.pdf on it. This is a paper for comparing KNN-WG and GCM data for future methods.

If you want one of the methods so you need one of the tools. However, if you want to compare two methods so you need both of them. For example, if you want to use this data in hydrology so you can generate data by using both methods and use it in hydrology then compare the result. But if you don't want to compare so you would select one of these methods.

By using our tools I can say you: Using GCM data is better than generating data by the KNN method but download GCM data and apply bias correction on it is a bit difficult that the KNN method.

Name: Hidden