Quantifying and mapping changes in hydrologic ecosystem services from a large magnitude wildfire in Shasta and Tehama counties, California
Year: 2014
Type: dissertation
Abstract: Ecosystem services are considered goods and services that sustain and \nenhance human life. Humans have degraded ecosystems due to land use or land \nmanagement practices favoring more profitable land use scenarios at the expense of other \nvaluable ecosystem services. Tools and models have emerged to quantify and value \necosystem services. InVEST is an open-source, spatially explicit, ecosystem service \nappraisal tool which has been widely-used. A review determined that only 18% of 153 \npublications on ecosystem services validated their results against observed data, and only \none third of the studies had any sound basis for their results. There are few studies which \nhave conducted a sensitivity analyses on InVEST model outputs. This study assessed \npredicted changes in the InVEST Hydropower Water Yield due to an episodic disturbance (e.g., fire), validated model results against observed data, and determined \nhow variation in spatial resolution (30, 100, 120, 250, 500, 1000, 4000 m) of input data \naffected model results. Results show that the model did not detect the fire event on water \nyield. The location of the fire is a more important element for producing a significant \nchange in water yield. After model results were calibrated against observed stream gauge \ndata, the model under-predicted by less than 1%. Annual or 30 year normal precipitation \ninputs do not significantly affect model results. Prior to calibration, normal precipitation \nresulted in a larger percentage of error 64% compared to annual data 44%. Calibration is \na crucial and understated step. Resolution of land cover and soil data does not always \ncreate significant variation in model results. InVEST may be driven more by parameter \nvalues (e.g., seasonality constant, root depth, and evapotranspiration coefficients) than \nthe resolution of soil and land cover data. These findings should assist InVEST users to \nmore accurately parameterize their models and think critically about the appropriate \nstudy area, data inputs, and accuracy in results allowing for a widespread assessment of \necosystem services so that they may sustainably managed.
Author: Daniel Larson
Language: English
Subfield: Global and Planetary Change
Field: Environmental Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W335687556