Journal
REMOTE SENSING OF ENVIRONMENT
Volume 252, Issue -, Pages -Publisher
ELSEVIER SCIENCE INC
DOI: 10.1016/j.rse.2020.112133
Keywords
Greenup onset date; Scale effect; HLS; VIIRS; Time series; Shrubland ecosystem components
Funding
- Strategic Priority Research Program of the Chinese Academy of Sciences [XDA19090112]
- National Natural Science Foundation of China [41571423]
Ask authors/readers for more resources
Shrublands cover 13% of the global land surface, but current satellite-derived land surface phenology products lack detection in some semi-arid areas. Research shows that green-up onset date shifts to earlier dates with increasing vegetation cover, becoming fully detectable at over 50% vegetation cover. This highlights the importance of considering vegetation cover when monitoring and assessing phenological dynamics in shrubland ecosystems.
Shrublands occupy about 13% of the global land surface, contain about one-third of the biodiversity, store about half of the global terrestrial carbon, and provide many ecosystem services to a large amount of world's human population and livestock. Because phenology is a sensitive indicator of the response of shrubland ecosystems to climate change, the alteration of ecosystems following species invasions, and the dynamics of shrubland ecosystem function, biodiversity, and carbon budgets, it is critical to monitor and assess phenological dynamics in shrubland ecosystems. However, most current land surface phenology (LSP) products derived from satellite data do not provide phenology detections in some semiarid shrublands where the amplitude of seasonal vegetation greenness is small. Therefore, we investigated the LSP detection using multiple spatial resolution satellite data and examined the impacts of spatial scales and shrubland ecosystem components (shrub and herb cover) on LSP detections over the western United States. Specifically, greenup onset date (GUD) in shrublands was detected from 30 m Harmonized Landsat and Sentinel-2 (HLS) data and 500 m Visible Infrared Imaging Radiometer Suite (VIIRS) data to quantify scale effects. The GUD spatial patterns were explored with 30 m pixel variations in shrubland ecosystem components. The results show that GUD values varied with percent vegetation cover and shifted to earlier dates with increasing vegetation cover, demonstrating that satellite observations were not able to capture greenup onset until a threshold of green vegetation cover is reached. GUD was mostly undetectable from both HLS and VIIRS pixels with vegetation cover less than 10% and became fully detectable with vegetation covers larger than 50%. Similarly, the differences of GUD between HLS and VIIRS detections also decreased with increased vegetation cover. As a result of high shrubland heterogeneity, GUD from 30 m HLS pixels could be partially detected within a 500 m pixel despite GUD being undetectable from VIIRS time series. Moreover, vegetation cover heterogeneity also made it difficult for GUD at 30 m to be aggregated to coarse scales (such as to 500 m VIIRS pixels). These findings have significant implications to the detection and characterization of shrubland LSP responses to environmental and climate changes.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available