Mason Archival Repository Service

Remote Sensing Techniques for Detecting Vegetation Phenology

Show simple item record

dc.contributor.advisor Qu, John J.
dc.contributor.author Li, Min
dc.creator Li, Min
dc.date 2010-07-27
dc.date.accessioned 2010-11-02T14:46:34Z
dc.date.available NO_RESTRICTION en_US
dc.date.available 2010-11-02T14:46:34Z
dc.date.issued 2010-11-02
dc.identifier.uri https://hdl.handle.net/1920/6027
dc.description.abstract Vegetation phenology describing the seasonal cycle of plants is currently one of the main concerns in the study of climate change and carbon balance estimation in ecosystems. In this study, we focus on vegetation phenology observed at landscape level. Phenology at the landscape scale poses challenges to observers because of its complexity, and it often generates confusion among observers because observers may use different approaches. Remote sensing techniques, which can capture canopy reflectance, allow vegetation photosynthetic capacity to be assessed, and provide the potential to move from plant specific observations to complete, continuous expressions of phenological patterns on the landscape. In this study, an improved satellite-based approach for detecting vegetation phenology was developed and the analysis of this satellite-derived phenology expresses an evident spatial pattern along latitude and elevation. Climate regulation of vegetation phenology shows that the vegetation phenological phases can be modeled using the annual mean Land Surface Temperature (LST). Two types of ecoregion-based models were established to compare the results with satellite-derived greenup onset dates. Comparing the satellite-based predictions with ground measurements demonstrated that the Normalized Difference Vegetation Index (NDVI) and the Enhanced Vegetation Index (EVI) are efficient at evaluating the greenup onset and dormancy onset respectively. The spatial analysis of satellite-derived phenological phases shows that a greenup wave is progressing northward in latitude and upward in elevation. The rate of change is about two days per degree latitude or per 100-meter elevation. The dormancy begins in the north and at higher elevation and then progresses southward in latitude and downward in elevation. The rate of change is about two days per degree latitude and one day per 100-meter elevation. The interannual variability of vegetation phenology is also identified by satellite measurements. The interannual variability of greenup onset is evident at higher latitudes (45-500N), while the interannual variability of the dormancy onset is larger at middle (35-450N) and lower latitudes (30-350N). The high goodness of fit (>0.8) indicates that the model based on annual mean LST predicts the average timing of vegetation phenological events successfully. As the annual mean LST rises, the average timing of greenup onset begins earlier, and the growing length is prolonged. The effect of global warming on vegetation greenup onset is evaluated by the thermal-chilling model. The results show that the thermal-chilling models can explain more than 80% of the variation in the Growing Degree-Days (GDD) required for greenup onset. Global warming may advance forest greenup onset when the chilling requirements are far exceeded and may delay greenup onset when the chilling requirements are nearly exactly sufficient. Finally, the ecoregion-based models have been established to simulate greenup onset dates and the results are compared with satellite-derived measurements. The greenup onset dates for 90% of the habitat-controlled ecoregions and 80% of temperature-controlled ecoregions are simulated within 10 days of the satellite derived greenup onset. The satellite-derived vegetation phenology is globally applicable. It is capable of identifying phenological behavior characterized by multiple growth and senescence periods. Remote sensing-based analysis provides a promising approach for the quantification of ecosystem-level response climate change, which is an important complement to species-level studies.
dc.language.iso en_US en_US
dc.subject vegetation phenology en_US
dc.subject remote sensing en_US
dc.subject MODIS en_US
dc.subject climate impact en_US
dc.subject ecoregion en_US
dc.subject temperature en_US
dc.title Remote Sensing Techniques for Detecting Vegetation Phenology en_US
dc.type Dissertation en
thesis.degree.name Doctor of Philosophy Geography and Geoinformation Science en_US
thesis.degree.level Doctoral en
thesis.degree.discipline Geography and Geoinformation Science en
thesis.degree.grantor George Mason University en


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search MARS


Browse

My Account

Statistics