Back to search

Boston University

Improving ecological forecasts using model and data constraints

Abstract

dc:description.abstract

Terrestrial ecosystems are essential to human well-being, but their future remains highly uncertain, as evidenced by the huge disparities in model projections of the land carbon sink. The existence of these disparities despite the recent explosion of novel data streams, including the TRY plant traits database, the Landsat archive, and global eddy covariance tower networks, suggests that these data streams are not being utilized to their full potential by the terrestrial ecosystem modeling community. Therefore, the overarching objective of my dissertation is to identify how these various data streams can be used to improve the precision of model predictions by constraining model parameters. In chapter 1, I use a hierarchical multivariate meta-analysis of the TRY database to assess the dependence of trait correlations on ecological scale and evaluate the utility of these correlations for constraining ecosystem model parameters. I find that global trait correlations are generally consistent within plant functional types, and leveraging the multivariate trait space is an effective way to constrain trait estimates for data-limited traits and plant functional types. My next two chapters assess the ability to measure traits using remote sensing by exploring the links between leaf traits and reflectance spectra. In chapter 2, I introduce a method for estimating traits from spectra via radiative transfer model inversion. I then use this approach to show that although the precise location, width, and quantity of spectral bands significantly affects trait retrieval accuracy, a wide range of sensor configurations are capable of providing trait information. In chapter 3, I apply this approach to a large database of leaf spectra to show that traits vary as much within as across species, and much more across species within a functional type than across functional types. Finally, in chapter 4, I synthesize the findings of the previous chapters to calibrate a vegetation model's representation of canopy radiative transfer against observed remotely-sensed surface reflectance. Although the calibration successfully constrained canopy structural parameters, I identify issues with model representations of wood and soil reflectance that inhibit its ability to accurately reproduce remote sensing observations.

Degree

thesis:*
Name thesis:degree_name
Doctor of Philosophy
Level thesis:degree_level
doctoral
Discipline thesis:degree_discipline
Earth & Environment
Grantor
Boston University
Year dc:date.issued
2018

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Shiklomanov, Alexey Nikolaevich
Advisor dc:contributor.advisor
  • Dietze, Michael C.

Subjects

dc:subject × 5

Rights

dc:rights
Statement dc:rights
  • Attribution 4.0 International
Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/2144/32714

Chain of custody

source
Harvested from
Boston University
Base URL
open.bu.edu/oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Shiklomanov, Alexey Nikolaevich. Improving ecological forecasts using model and data constraints. doctoral thesis, Boston University, 2018. https://hdl.handle.net/2144/32714