OverviewΒΆ
Introduction
This first recipe covers essential features of gospl in the form of interactive Jupyter notebooks and Python scripts. Over the course of this recipe, you will learn:
data structure used in
gosplinput file,how to generate simple initial conditions like topography, tectonic and precipitation to force a simulation,
how to extract some of the outputs from the model to visualise them in Jupyter notebooks using multiple approaches/libraries,
how to export
gosploutputs (hdf5) in other common formats such as netcdf, geotiff, ZMap, andhow to extract information relative to drainge basins and rivers longitudinal profiles.
The simulation proposed here uses data freely available from the web and will run without additional dataset. The dataset are imported in the jupyter notebooks from the following THREDDS servers:
Topography file from NOAA ETOPO5,
Precipitation file from the CPC collection.
In this example, we show how to use the global model to run a simulation for a particular continental-scale region here the Gulf of Mexico. This is done by building an unstructured spherical mesh with high resolution (approx. 25 km) around North America and a coarse (>1000 km) elsewhere.
Note
The recipe should take around 20/30 minutes to do on a single CPU (actually the simulation itself runs in less than 5 minutes but you will need time to understand how to build the input files and perform the provided post-processing analysis). Obviously, gospl can be ran at much higher resolution (<10 km) globally but will take much more time!
The recipe is actually a coarser version of the following simulation which takes a bit longer to run but has been built using the exact same notebooks.
