A virtual globe for land degradation assessment and hydrology

Assessment of environmental impacts and natural hazards is increasingly required on a global or continental scale. Analyses at this scale not only transcend political boundaries, but also require input data with global or continental coverage. The high resolution numerical models used in analyzing soil erosion, storm damage, forest fires, landslides, or hydrological hazards often operate at a much smaller scale, requiring detailed local input data which is not available globally or in data-scarce regions. Another problem is that global data is often static, while impact and risk assessment studies require dynamic spatio-temporal data generated from specific user-defined scenarios. There are also obstacles of a technical nature: researchers refactoring models to seamlessly operate on larger areas in other parts of the world face restrictions on computational resources and difficulties in working with large global datasets. This has made access to global and continental scale modeling difficult to all but experienced modellers. To address these problems we have developed a web-based platform which allows both researchers as well as policy and decision makers to configure and run high resolution (±100m) environmental models nearly anywhere in the world. Various global datasets have been reprojected and resampled to be used directly as model inputs. Currently available to modelers are the SRTM elevation model, global monthly vegetation data from MODIS, land use classifications, historical climate data, soil information, population density, and near real time weather forecasts. The platform uses the PCRaster-Python modeling framework to process spatio-temporal models in real time; any attribute maps, timeseries data, or model scenarios can be instantly shared or downloaded via a web mapping service (WMS).

The current system only includes a limited number of highly simplified models (snow accumulation and melt, flooding). The aim of this topic is to add more sophisticated model(s) of a particular process of your choice (for instance water erosion, wind erosion, crop growth). You will need to address, for instance, the following research questions:

  • What is the optimal model structure (i.e. inputs, equations) given the limited data availability and the possibly large spatial variation in processes at the global scale?

  • What is the quality of the model results (compare model results with observations)?

  • What is the performance of the model compared to other (local) model studies that used more detailed data (model intercomparison)?

This is an interesting topic if you like to formulate concepts for process-based models of the environment, and if you are not afraid of programming such models. As the current system uses PCRaster Python, models need to be developed in PCRaster Python. After model development, it is relatively straightforward to upload them to the virtual globe software.

The virtual globe is available at http://virtualglobe.geo.uu.nl. Email Derek Karssenberg (d.karssenberg@uu.nl) to retrieve a login and password.

Supervision: Dr Derek Karssenberg (Utrecht University), in cooperation with colleagues (depends on application) and the PCRaster software development team.

Location: Utrecht University

Period: to be determined

Number of students: 1 - 4

Program/track: Earth Surface Hydrology or Natural Hazards and Earth Observation

Prerequisites: courses in spatio-temporal modelling (preferably PCRaster Python), hydrology, geomorphology, and/or natural hazards (content of project can be adjusted to your background)

Contact/info: Derek Karssenberg (d.karssenberg@uu.nl)