UNRESP is a research project working on building resilience to an environmental pollution hazard caused by persistent volcanic emissions. The volcanic emissions are termed vumo.

UNRESP is working on Masaya volcano in Nicaragua, one of the biggest volcanic polluters in the world. Very little is known about the dispersion and levels of the air pollution caused by the vumo, and there no established procedures for communicating the vumo hazard between the local authorities and the public.


Create whole package forecasting tool for volcanic emissions Main Tool CALPUFF


Setting a constant point source emission of Sulphides and Sulphates from Masaya Volcano, daily NAM Data is used to drive a light weight dispersion model (CALPUFF) to forecast over the next 48 hours qualitative concentrations of SO2 and SO4. The idea being this could be combined or constrained with observational data and ran in a matter of minutes to give a quick rough estimates of air quality for the following day

unresp gif

Portability and documentation

This tool has been designed to work on any linux or Mac OS (requires bash, FORTRAN and python) and to work seamlessly to install compile and run the model as well as visualise the output.

We have tested running this set up on a variety of different machines across the world with success, running the model set up with minimal effort.

The end goal is to use this a INITER in Nicaragua to allow local volcanologists to use this as a supplementary tool.

A full repository of code and information is hosted on github .


Along side the modelling code are a set of python tools to automatically generate images and animations of the concentration field for each day. These can also be used as stand alone tools and have been used by our colleagues with ease.

A simple html and JavaScript framework to host or simply browse the images is also provided and can be launched in its entirety on any web server even a simple HTTP python server.

uresp gif

Skills Used

CALPUFF – FORTRAN – bash – Python – js – css – html – wgrib – eccodes