Plotting and Post-processing

Overview

Users rely on a diversity of platforms and approaches for interacting with and visualising GLM output. The user community has developed a range of scripting and plotting tools. This page is always a bit behind the numerous initiatives going on, but look below to get started with different types of analysis tools.

Options include:

    Plotting:

  • In-built plotting: Depth-profile contours are built into the distributed model binary
  • Excel: Plot essential time-series of model outputs from the generated csv files
  • R: Create high-quality model outputs using the GLMr tools
  • MATLAB: Create high-quality model outputs using the GLMm scripts
  • Processing and Analysis:

  • Excel: Use *.csv outputs to look at water and heat balance, depth-specific lake properties, and outflow properties
  • R: Use the GLMr and associated tools such as LakeAnylseR for advanced plotting and analysis
  • MATLAB: Use the GLMm and associated scripts for advanced plotting, analysis and nutrient budgeting etc
  • Other:

    • PEST: Examples of GLM use within the PEST uncertainty analysis platform are available
    • glmGUI: R-based GUI
    • GLM-i: New GUI under development
    • MCMC: Calibration with Bayesian uncertainty assessment
    • GRAPLEr: R-based distributed computing tool for managing large volumes of GLM simulations

General lake information: lake.csv

The model automatically creates a file called lake.csv, that contains information about the lake volume balance, surface heat balance and other lake-scale variables that may be of interest.

Creating a custom time-series output file: custom_depth.csv

The model can be configured to produce a time-series of any chosen variables at a custom depth. This is useful for plotting model output in Excel, R or MATLAB.

Using the in-built time-depth contour plots: plots.nml

If the xdisp command line argument is invoked, then a plot window will open and display results as the model is running. This plot window is entirely customisable, and this is done via the plots.nml text file. This file allows setting of the window size, number of plots, variables to plot, and color limits. Variables that can be plotted include: temp, salt, rad and any FABM state or diagnostic variable.

Plotting with R: glmtools

The model can be Managed within R via the GLMr and glmtools packages. Click here for the glmtools tutorial (v3.0: Nov 2019).

Plotting with R: glmGUI

The model has a GUI available within the R environment. See this paper for the glmGUI overview.

Analysing the results with LakeAnalyser

The model results area able to be analysed using the Lake Analyzer program developed by Read et al. (2012). In particular modellers can use this tool to compare their simulations against metrics such as thermocline depth, lake number and Schmidt stability etc. Refer to lake results for examples. MATLAB scripts to use GLM with LA are available on request.

Figure : Overview of Lake Analyser Output

Calibrating the model with the MCMC Tool

The model may be run with the MCMC Toolbox For Matlab in order to quantify uncertainty in model predictions. Users may run without MATLAB using a pre-compiled MCMC executable package: runMCMC

Note users will need the MATLAB R2012a MCR installed MATLAB Compiler Runtime.

Figure : MCMC Output

Style Switcher

Predefined Colors

Layout Style

Background Image

Background Color

Reset