Mapping and spatial-temporal modeling of Bromus tectorum invasion in Central Utah

Update Item Information
Publication Type dissertation
School or College College of Social & Behavioral Science
Department Geography
Author Jin, Zhenyu
Title Mapping and spatial-temporal modeling of Bromus tectorum invasion in Central Utah
Date 2011-11
Description Cheatgrass, or Downy Brome, is an exotic winter annual weed native to the Mediterranean region. Since its introduction to the U.S., it has become a significant weed and aggressive invader of sagebrush, pinion-juniper, and other shrub communities, where it can completely out-compete native grasses and shrubs. In this research, remotely sensed data combined with field collected data are used to investigate the distribution of the cheatgrass in Central Utah, to characterize the trend of the NDVI time-series of cheatgrass, and to construct a spatially explicit populationbased model to simulate the spatial-temporal dynamics of the cheatgrass. This research proposes a method for mapping the canopy closure of invasive species using remotely sensed data acquired at different dates. Different invasive species have their own distinguished phenologies and the satellite images in different dates could be used to capture the phenology. The results of cheatgrass abundance prediction have a good fit with the field data for both linear regression and regression tree models, although the regression tree model has better performance than the linear regression model. To characterize the trend of NDVI time-series of cheatgrass, a novel smoothing algorithm named RMMEH is presented in this research to overcome some drawbacks of many other algorithms. By comparing the performance of RMMEH in smoothing a 16-day composite of the MODIS NDVI time-series with that of two other methods, which are the 4253EH, twice and the MVI, we have found that RMMEH not only keeps the original valid NDVI points, but also effectively removes the spurious spikes. The reconstructed NDVI time-series of different land covers are of higher quality and have smoother temporal trend. To simulate the spatial-temporal dynamics of cheatgrass, a spatially explicit population-based model is built applying remotely sensed data. The comparison between the model output and the ground truth of cheatgrass closure demonstrates that the model could successfully simulate the spatial-temporal dynamics of cheatgrass in a simple cheatgrass-dominant environment. The simulation of the functional response of different prescribed fire rates also shows that this model is helpful to answer management questions like, "What are the effects of prescribed fire to invasive species?" It demonstrates that a medium fire rate of 10% can successfully prevent cheatgrass invasion.
Type Text
Publisher University of Utah
Subject Cheatgrass; NDVI time-series; Remote sensing; Spatial-temporal modeling; Utah; Bromus tectorum
Dissertation Institution University of Utah
Dissertation Name Doctor of Philosophy
Language eng
Rights Management Copyright © Zhenyu Jin 2011
Format Medium application/pdf
Format Extent 5,152,189 bytes
Identifier us-etd3,81487
Source Original housed in Marriott Library Special Collections, QK3.5 2011 .J56
ARK ark:/87278/s6k64ztq
Setname ir_etd
ID 194754
Reference URL