Hydraville!

This repository contains a complete example Pywr model. The fictional city of Hydraville is located in a river basin within a river basin with significant water resource challenges. A Pywr model has been built to aid the management and planning of those challenges.

This model serves as a training and tutorial example for water resource planning and decision making under uncertainty.

Overview

Installation

Obtaining the data

INSERT INSTRUCTIONS TO DOWNLOAD PREPROCESSED DATA FROM S3

Basic usage

The basic usage of this repository requires one to first generate a Pywr input file. The repository allows many configurations of the same core models to be generated. To begin we create the most basic water only model derived from the "water-simple" configuration.

We recommend the user create an analysis subfolder to save the generated models and their outputs in.

hydraville create analysis/water-only.json --model water-simple

Once the input file is generated we can run the model:

hydraville run analysis/water-only.json

UKCP09 data pre-processing

Importing in to Hydra

Data processing

This repository also contains code for pre-processing and sub-sampling UKCP09 weather data, and subsequently running that data through a series of hydrological models written in Catchmod (using pycatchmod).

UKCP09 Pre-processing

The output of a UKCP09 Weather Generator run is assumed to be saved in the data/ directory. This output should consist of a series of ZIP archives containing CSV files. The following command will pre-process that data by exporting it in to a single HDF5 file.

hydraville ukcp09 preprocess data/ data/ukcp09_full.h5

UKCP09 Sub-sampling

The full output of a UKCP09 Weather Generator run contains many scenarios. For our modelling purposes we may want to sub-sample this set to a more managable number. The following command will sub-sample the previously created combined archive. Each sub-sample is saved in a separate file with a suffix (e.g. data/ukcp09_sub010.h5)

hydraville ukcp09 subsample data/ukcp09_full.h5 data/ukcp09.h5

Running catchmod

Once the sub-sampled data has been created. The hydrological flows can be generated using Catchmod. The catchmod output is saved in to formats. The first is with a dataframe for each of the UKCP09 files (analagous to the UKCP09 files created above). The second groups all the flows for each of the catchmod models into a single dataframe with a column for each UKCP09 scenario.

Please note that running catchmod takes a long time for all scenarios.

hydraville catchmod run data/ukcp09.h5 data/catchmod_flows.h5 data/catchmod_flows_by_catchment.h5