First Advisor

Samantha Hartzell

Term of Graduation

Summer 2022

Date of Publication

7-12-2022

Document Type

Thesis

Degree Name

Master of Science (M.S.) in Civil & Environmental Engineering

Department

Civil and Environmental Engineering

Language

English

Subjects

Evapotranspiration, Machine learning

DOI

10.15760/etd.8040

Physical Description

1 online resource (x, 112 pages)

Abstract

Machine learning has been used as a tool to model transpiration for individual sites, but few models are capable of generalizing to new locations without calibration to site data. Using the global SAPFLUXNET database, 95 tree sap flow data sites were grouped using three clustering strategies: by biome, by tree functional type, and through use of a k-means unsupervised clustering algorithm. Two supervised machine learning algorithms, a random forest algorithm and a neural network algorithm, were used to build machine learning models that predicted transpiration for each cluster. The performance and feature importance in each model were analyzed and compared to produce (1) transpiration models that can generalize to new locations and (2) insights into the environmental variables that control each cluster. There was high variance in feature importance between clusters, yet many clusters showed good performance, indicating that key predictors of transpiration vary by climate. High performance was achieved for clusters with enough data to define the group, but not for clusters so large that the machine learning algorithm was unable to define an underlying pattern. High-performing models achieved R2 values to measurement data in the range of 0.74 to 0.97. Water-limited climates tended to be more controlled by soil water content, whereas climates with high mean annual temperature tended to be more controlled by solar radiation and less dependent on air temperature. By defining which cluster a site fits into, these novel generalized models can be used to predict transpiration and the variables that control it. The predictions provide climate-specific insights into how forests respond to their environment.

Rights

© 2022 Morgan Tholl

In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).

Persistent Identifier

https://archives.pdx.edu/ds/psu/38765

Download

Share

COinS