Modelling the hydrological impact of wildfire in MESO-scale catchments of the Garhwal Himalaya
Student name: Mr Sayan Roy
Guide: Dr Joachim Michael Schmerbeck
Year of completion: 2015
Host Organisation: TERI University and University of Augsburg, Germany
Supervisor (Host Organisation): Prof Peter Fiener
Abstract: This study makes an attempt to model the hydrological consequences of wildfire occurrence in
the region of the Garhwal Himalaya. A suitable model, the Automated Geospatial Watershed
assessment tool (AGWA) is chosen to simulate different vegetation and fire scenarios. The model
is parametrized for land cover using primary vegetation studies plot level data. KINEROS is
chosen as the rainfall simulation model, and a hypothetical storm event is set up based on
nearby rain gauge data. Parameters such as Manning’s roughness and cover are modified based
on assumptions, rules and where available, cover data for post-wildfire conditions. Simulations
are subsequently run with pre-fire as well as post-fire scenarios, with two cases of wildfire
prescription to the catchment; indiscriminate and patchy occurrence of wildfire, the latter based
on fire frequency data prepared from LANDSAT. Model outputs are visualized in different ways,
spatially distributing outflow (runoff) across sub-catchments, as well as sediment yield
distribution. Hydrographs are generated for a specific sub-catchment and compared and
evaluate across different fire scenarios. It is found that the model is able to produce distinct prefire
hydrological outputs by virtue of the cover data. However, post-fire simulation reveals
AGWA to make no distinction between different vegetation cover, wildfire always resulting in the
same effect on runoff and sediment outputs. In general, the observable trends are that runoff and
sediment yield maximum values increase significantly when the entire catchment is burned. In
the case of only parts of the catchment being affected by wildfire, the resulting sediment and
runoff values fall somewhere in the middle of pre-fire and outright post-fire simulations. Spatial
distributions produced by AGWA reveal interesting findings, however, better data sets need to be
incorporated into model to improve accuracy of simulation.
