Skip to content

LSDtopotools/MuddPILE

Repository files navigation

MuddPILE License GPL%20v3 blue 104057321

MuddPILE: The Parsimonious Integrated Landscape Evolution model

Welcome to MuddPILE. This is a simple landscape evolution model built into LSDTopoTools. The authors are Simon Mudd, James Jenkinson, Fiona Clubb and Declan Valters. It might seem narcissistic to call it MuddPILE given that it was a collaborative project but you must admit that calling a landscape evolution model "MuddPILE" is slightly amusing. If you do not think it amusing at all and belive that science should be very serious then you probably shouldn’t ever try to work with Simon Mudd, who is writing this readme in the third person.

What is MuddPILE?

MuddPILE is a very simple landscape evolution model that has fluvial erosion modelled with the stream power law (E = K Am Sn) and hillslope evolution modelled with a nonlinear sediment flux law, e.g., Roering et al., 1999, JGR. It uses the FASTSCAPE algorithm to compute fluvial erosion, and I (SMM) cooked up a relatively fast implicit method to solve the hillslope equations since they are nasty and nonlinear. You can force the model with spatially heterogeneous uplift and erodibility. It runs fairly quickly. So if your goal is just to generate some fairly simple landscapes with some fairly simple rules fairly quickly this is the model for you.

The first appearance of this code was in a 2017 ESPL paper, but we have now placed it online in an effort to ensure reproducibility. We hope you find the model useful.

Compile instructions

MuddPILE is written in c++ and distributed as source code so you need to compile it. Instructions are below.

Prerequisites

Firstly, this code was developed in Linux and easier to install in Linux systems than other systems. Do you not run Linux? Nae problem.

If you want a full explanation of getting a Docker container, please have a look at our LSDTopoTools installation documentation. MuddPILE has a slightly different installation than other LSDTopoTools programs, however, which we describe below.

  1. First you need all the GNU c++ tools and make. But basically every Linux operating system has this so don’t worry about it. Our docker containers have them pre-installed.

  2. Second, you need something called FFTW. If you already installed [LSDTopoTools] it is already on your computer. If not, install it like this (this is for Ubuntu, if you’ve not got apt-get then try yum):

    $ sudo apt-get install -y libfftw3-dev
  3. Third, you need to decompress the boost and mtl4 libraries. We have packaged these up for you so our MuddPILE distribution is self-contained. To unzip, go into the boost_mtl_minimal directory and unzip using tar:

    $ tar -xvzf boost_mtl_minimal.tar.gz
  4. Fourth, go into the /src/driver_functions/ directory and make the program:

    $ make -f MuddPILEdriver.make
  5. The result will be a program called MuddPILEdriver.out: this is the model! You can now run landscape evolution runs. More documentation can be found at https://lsdtopotools.github.io/LSDTT_documentation/LSDTT_MuddPILE.html