Monitoring Global Forest Land-Use and Change

<p>Earth’s forests contain nearly three-fourths of the World’s floral and faunal diversity, function as a large carbon sink capable of mitigating the effects of global climate change, affect local and regional physical and chemical cycles and provide wood and non-wood products. However, humans are now capable of modifying their environment in ways more impactful and at rates faster than at any other time in history. Consistent and comparable estimates of global forest land-use and change are critical for monitoring human impacts on the Earth system. International treaties and reporting requirements aimed at safeguarding the delivery of forest-related ecosystem services depend on such estimates for measuring progress against their stated goals. Many existing studies have estimated tree cover and change at a variety of spatial scales from local to global. However, this existing research focuses largely on land cover classification, but generally lacks ecological context for estimating true human land use. The objective of this dissertation is to address this gap by exploring how forest land use can be mapped and monitored using medium spatial resolution optical satellite imagery in order to estimate forest land use change over time for large geographic areas. First, the effects of clouds, cloud shadows and missing data were analyzed to determine the amount of moderate spatial resolution, optical satellite data needed to detect and map land cover changes over large, spatially continuous areas on frequent time intervals. Second, an alternative method to spatially exhaustive mapping was developed and tested for estimating land cover and land use change globally employing object-based image analysis and a sample-based estimation approach. The method facilitated expert human intervention to identify true land use change in an operational way. Finally, these methods were applied to a globally distributed sample of remotely sensed data for the time periods 1990, 2000 and 2005. The results of this research produced the first consistent and comparable global time-series dataset of forest land-use estimates.</p>

Name thesis:degree_name

Doctor of Philosophy (PhD)

Level thesis:degree_level

Dissertation - Open Access

Discipline thesis:degree_discipline

Geospatial Science and Engineering

Year dc:date.available

2016

Author dc:creator

• Lindquist, Erik J.

Contributors dc:contributor

• Matthew Hansen

• change
• forest
• global
• landsat
• land-use
• Geographic Information Sciences
• Physical and Environmental Geography
• Remote Sensing

Licence

• http://rightsstatements.org/vocab/InC/1.0/

Language dc:language

en

Repository record dc:identifier

https://openprairie.sdstate.edu/etd/997

OAI identifier oai:identifier

oai:openprairie.sdstate.edu:etd-1997