Downloading and conversion of data
Remote sensing and modelled datasets are often distributed in image or swath based formats. These formats are not ready to be used for efficient time series analysis. They have to be converted into a time series representation. The dataset packages that belong to Poets° offer functionality to download and convert the data into a time series format. Downloading can of course also be performed manually but using the included command line scripts stores the data in the correct folder structure which makes it easier to use the commands for time series conversion.
The chosen format for the time series is one of the representations defined in the CF Conventions. The exact format is described in the documentation of each package. Please find the detailed information about the conversion at the following locations.
ASCAT data is already available in a timeseries format so no conversion is necessary.
Reading of time series data
After conversion the time series are all accessible under the same API.
# Example of reading GLDAS time series data
from gldas.interface import GLDASTs
ds = GLDASTs(ts_path)
# read_ts takes either lon, lat coordinates or a grid point indices.
# and returns a pandas.DataFrame
ts = ds.read_ts(45, 15)
Detailed instructions for reading the data can be found in the documentation of each dataset package linked above.
Using the data in the pytesmo validation framework
The pytesmo validation framework implements the following processing modeled:
More detailed information about how the framework works can be found in the pytesmo documentation. This example currently shows a validation of ISMN data and the ASCAT products. Other examples are in preparation.
The framework can be used together with the datasets supported by Poets° but it is also possible to write compatible interfaces to datasets you might already have. The basic requirement is that the dataset can be accessed by a function that takes the two arguments longitude and latitude and returns a pandas.DataFrame.