|GERVAIS, N. - Albany State University
|BUYANTUEV, A. - Albany State University
Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/17/2017
Publication Date: 1/23/2017
Citation: Gervais, N., Buyantuev, A., Gao, F.N. 2017. Effects of the urban built-up environment on plant phenology: high-resolution analysis with fused Landsat and MODIS data. Remote Sensing of Environment. doi:10.3390/rs9010099.
Interpretive Summary: Urban land cover is usually significantly warmer than its surrounding rural areas. The relationship between urban intensity and plant phenology has been observed. However, these relationships are not well quantified due to the high heterogeneity of urban landscapes. This study investigates the differences of plant phenology for urban and rural areas using the high temporal and spatial resolution remote sensing data generated through data fusion approach. The results show strong evidence that the start of the season is earlier in the urban area, contributing to a longer length of season, as an example within the vicinity of Ogden, UT. Understanding the effects that the Urban Heat Island has on plant phenology is important in modeling ecological impacts of expanding cities and land cover changes. This type of information is useful for local farmers and land operators, including local food producers.
Technical Abstract: Understanding the effects that the Urban Heat Island (UHI) has on plant phenology is important in predicting ecological impacts of expanding cities and also the projected increase in global temperatures. However, observing these effects is often limited by the underlying methods to monitor phenological events. Generally, one can either have a small sample of in situ measurements or use satellite data to observe large areas of land surface phenology (LSP) where a tradeoff exists among platforms with some allowing better temporal resolution to pick up discrete events and others possessing the spatial resolution appropriate for observing heterogeneous landscapes, such as urban areas. To overcome these limitations, we applied the Spatial and Temporal Adaptive Reflectance Model (STARFM) to fuse Landsat and MODIS data and derived high temporal and high spatial resolution synthetic Normalized Difference Vegetation Index (NDVI) time-series of imagery to identify the dates of the start of the growing season (SOS), end of the season (EOS), and the length of the season (LOS). The results were compared between the urban and exurban developed areas within the vicinity of Ogden, UT. The results show strong evidence that the SOS is earlier in the urban area, contributing to a longer LOS. However, statistical results were inconsistent in finding differences in the EOS. Although there was strong evidence that STARFM was able to produce images capable of capturing the UHI effect on phenology, we recommend that future work help to evaluate the value of this method over using just one data source.