Year of Award


Document Type


Degree Type

Master of Science (MS)

Degree Name


Department or School/College

Department of Geography

Committee Co-chair

David Shively, Zachary Holden

Commitee Members

Andrew Larson


Gradient Model, Remote Sensing, Species Distribution Model, Western Larch


University of Montana


Western larch (Larix occidentalis Nutt.) is one of three native North American larch species, it occupies the mountainous regions of northwestern North America, and it is a deciduous conifer. Western larch is among the most ecologically and economically important conifer tree species in the northern Rockies region. This study explores the viability of mapping western larch via the analysis of multi-temporal Landsat imagery and gradient modeling. Larch presence and abundance data from 300 field plots correlated with Normalized Difference Vegetation Index seasonal change (NDVIsc) explains 46% of the variability in larch basal area. Multivariate models built with NDVIsc and additional climatic and topographic variables only slightly improved the models. These satellite imagery based models suggest that western larch tends to occur primarily on shaded, north-facing slopes within the study area. This analysis was contrasted with a gradient modeling approach using data from 4800 Forest Inventory and Analysis plots and a suite of fine scale (30-60 m) topographic and climatic data as predictors. These models correctly predicted larch presence with error rates of less than 20%. Presence or absence of western larch was found to be strongly dependent on minimum temperature and water balance variables (soil moisture deficit and actual evapotranspiration). Probability prediction rasters produced with these data also showed a noticeable northern aspect tendency. The accuracy of the remote sensing based models suggest that the method may be applied to other areas, and the output from both model types points to a strong relationship between larch presence and fine scale topographic and climatic factors, especially as they interact to affect soil moisture.



© Copyright 2013 Steven Joseph Touzel