Skip to content

Projection

The Projection page explores how streamflow might change under future climate scenarios. Using calibrated model parameters and climate model projections, you can assess potential impacts of climate change on water resources.

Overview

Climate projection simulates streamflow using:

  1. Calibrated model parameters from the Calibration page
  2. Future climate data from climate projections

Results show potential changes in streamflow seasonality and magnitude under different climate scenarios and time horizons.

Projection page

Prerequisites

Before using projections, you need:

  1. Calibrated parameters: Export from the Calibration page
  2. Projection data: Climate projections for your catchment (pre-loaded in HOLMES)

Projection Data Availability

Projection data is catchment-specific. Not all catchments have projection data available.

Importing Calibration Results

Step 1: Prepare Calibration

  1. Go to the Calibration page
  2. Calibrate a model for your catchment
  3. Click Export parameters to save the JSON file

Step 2: Import to Projection

On the Projection page:

  1. Click Import model parameters
  2. Select your calibration JSON file
  3. The calibration details appear in the table

Single Calibration

Unlike Simulation, the Projection page uses only one calibration at a time. Importing a new calibration replaces the existing one.

Calibration Results Table

The table displays the imported calibration:

Field Description
hydrological model GR4J, Bucket, etc.
catchment Must match available projection data
objective Calibration objective used
transformation Streamflow transformation
algorithm Manual or SCE
date start Calibration period start
date end Calibration period end
snow model CemaNeige or none
parameters Calibrated parameter values

Projection Settings

After importing a calibration, configure the projection:

Climate Model

Select the climate model that provides future climate data:

  • Different models have different assumptions and biases
  • Multiple models allow uncertainty assessment

Horizon

Select the future time period. Available horizons depend on the projection data for your catchment (e.g., H20, H50, H80).

Climate Scenario

Select the climate scenario. Available scenarios depend on the projection data for your catchment (e.g., RCP4.5, RCP8.5, REF).

Running a Projection

  1. Import a calibration result
  2. Select climate model, horizon, and scenario
  3. Click Run

A loading indicator shows while the projection runs (this may take longer than calibration/simulation due to the longer time series).

Projection results

Understanding Results

Projection Chart

The main chart shows the mean daily hydrograph (averaged across all years in the projection period):

  • X-axis: Month (January to December)
  • Y-axis: Streamflow
  • Light lines: Individual ensemble members
  • Bold line: Median across all members

This shows the expected seasonal pattern of streamflow under the selected scenario.

Chart Interactions

  • Zoom: Click and drag to select a time range
  • Reset: Double-click to return to full year view

Results Chart

The dot plot shows summary statistics for each ensemble member:

Metric Description
Winter min Minimum daily flow in January-March
Spring max Maximum daily flow in March-June
Summer min Minimum daily flow in May-October
Autumn max Maximum daily flow in September-December
Mean Annual mean streamflow

Each dot represents one ensemble member, showing the spread of projections.

Interpreting Projections

Ensemble Spread

The spread of ensemble members indicates uncertainty:

  • Narrow spread: Models agree on the projection
  • Wide spread: Significant uncertainty in the projection

Seasonal Changes

Look for changes in:

  • Peak timing: Has the spring freshet shifted earlier/later?
  • Peak magnitude: Are floods projected to increase/decrease?
  • Low flow timing: When do minimum flows occur?
  • Low flow magnitude: Are droughts projected to worsen?

Comparing Scenarios

Run projections with different settings to understand:

  • Horizon effect: How do projections change across different time horizons?
  • Scenario effect: How do different climate scenarios compare?
  • Model effect: Do different climate models agree?

Exporting Results

Click Export data to save:

  1. Projection timeseries (CSV): Daily mean streamflow by day of year

    date,01,02,03,median,model,horizon,scenario
2021-01-01,15.2,14.8,16.1,15.2,CSI,H50,RCP4.5
2021-01-02,14.9,14.5,15.8,14.9,CSI,H50,RCP4.5
...

  2. Projection results (CSV): Summary statistics per member

    member,winter_min,summer_min,spring_max,autumn_max,mean
01,5.2,2.1,125.3,45.6,22.4
02,4.8,1.9,118.7,42.3,21.1
...

Common Issues

No Projection Data

If the projection settings don't appear after importing calibration:

  • The catchment may not have projection data available
  • Try a different catchment with projection data

Long Run Times

Projections can take longer than calibration:

  • Climate model data spans many years
  • Multiple ensemble members are processed
  • Wait for the loading indicator to complete

Unexpected Results

If projections seem unrealistic:

  • Verify calibration quality first
  • Check that the correct catchment was calibrated
  • Consider if the model is appropriate for future conditions