Background

What's the deal with the name?

The name indicates that RASpuTIN can convert RASter files to Triangulated Irregular Networks (TINS). The pu in the middle just makes an awesome name.

Why is it developed?

Five years ago, one of Rasputin's authors was a main developer of an Open Source hybrid C++/Python framework for doing distributed hydrological modelling and forecasting, called SHyFT. It was clear then, that in order to get more precise forcing data into the models, information about slope, aspect and shades, in addition to height, was needed. However, it was not a prioritised task at the time. The idea of constructing a tool that can read raster DEMS and convert these into surface meshes is quite natural and has been done several times before. However, we wanted to have full control of the software stack, and utilise as much existing high-quality software as possible.

Where can Rasputin come in handy?

We are considering several application areas. Of course, we still have hydrological modelling, where we can use the information about slope, aspect and shade to locally adjust radiation forecasts and measurements in order to better resolve the energy balance in a watershed area. This should give better snow pack behaviour and hopefully enable a better understanding of when the snow will melt.

Another application is to use Rasputin to forecast avalanche dangers. Our current take on this is to read existing avalanche forecasts from an external source, and use the information about exposed elevations, aspects and slopes to colour the terrain based on known dangers.

Another application of Rasputin is to generate nice geometries that can be 3D-printed :)

Future work

There is a lot of work that needs to be done in order to do proper avalanche forecasting. First, a realistic physically based model for a snow pack needs to be implemented. Candidates here are SNOWPACK and Crocus. These models are one dimensional, but still computational expensive to solve for each face in a TIN. Second, as wind deposited snow is an extremely important factor for avalanche dangers, having a model that model local wind effects needs to be implemented. Then we need a model for how snow is moved by this wind field, and couple this to the snow pack model.

It would be cool to make a system that allows users of smart phones to see the avalanche dangers for the area surrounding the user, by using the GPS position and the orientation of the phone.

In hydrological modelling, Rasputin could automatically compute watershed areas for any given point, and also compute water routes in the terrain. Having a webgl output format, Rasputin can be used to visualise various geometrically distributed fields to guide the interpretation of model results from e.g. SHyFT.

Also, please see the Roadmap.