Remote sensing of vegetation phenology and snow-cover extent in Northern Colorado Plateau Network parks: status and trends 2010

Thoma DP. 2011. Remote sensing of vegetation phenology and snow-cover extent in Northern Colorado Plateau Network parks: status and trends 2010. Natural Resource Technical Report. NPS/NCPN/NRTR—2011/455. National Park Service, Natural Resource Stewardship and Science. Fort Collins, Colorado

Remote sensing of vegetation condition and snow extent provides information essential to understanding and managing change at broad scales by describing important aspects of ecological and hydrological processes that have cascading effects on other natural resources. The spatial and temporal pattern of satellite-derived NDVI (Normalized Difference Vegetation Index), or “greenness,” and vegetation is indicative of long-term climate conditions to which vegetation on the Colorado Plateau have adapted. NDVI also responds to short-term changes in weather and cycles annually with seasons. Changes in precipitation and temperature may cause changes in the spatial distribution of vegetation over time, and changes in NDVI may imply a response to changing climate or changing land condition brought about by human activities or natural disturbance. For this report, MODIS snow-cover extent and NDVI image products from 2000 through 2009 were processed and analyzed for 16 national park units on the Northern Colorado Plateau. Significant NDVI image-processing steps involved masking for clouds, aerosols, and snow, then smoothing annual curves with weighted least-squares regression. NDVI summary statistics were determined at the park scale, as were weather station data where they were available. NDVI time-series data were used to calculate phenology metrics with Timesat software. Areal snow-cover extent was determined as a percentage of park area covered by snow at eight-day intervals. Temporal trend analysis at the park scale was determined via linear regression for phenology and snow-cover metrics. Relationships between precipitation abundance and timing, as well as temperature, with NDVI were apparent for most parks where NDVI response mirrored, but lagged precipitation by 2–4 weeks, depending on season. The importance of winter soil-moisture recharge was made apparent by below-normal NDVI in years of lower-than-normal winter precipitation. There were clear spatial trends in the 2009 monsoon NDVI anomaly, both within parks and across the Colorado Plateau, likely related to monsoon storms tracking a more northerly direction than in previous years. Trends in NDVI and snow cover (p-value <0.10) at the park scale were inconsistent in direction and generally weak, which is not unexpected for a short, 10-year record across disparate parts of the Colorado Plateau landscape—even if significant change may have occurred. In this report, statistical significance does not imply biological or physical significance. This is because “step-trends” can occur, in which several years of consistent values might be be followed by a “jump” to a different, but equally consistent value for several years. Step trends are expected with quasi-periodic fluctuations in climate that tend to cycle at irregular intervals of between two and seven years. The most frequent trends in phenology were for (1) end-of-season date, which came earlier at three parks (Capitol Reef National Park, Colorado National Monument, and Hovenweep National Monument) and later at two parks (Golden Spike National Historic Site and Dinosaur National Monument); and (2) day of peak NDVI, which came later at three parks (Arches National Park, Black Canyon of the Gunnison National Park, and Dinosaur National Monument) and earlier at one park (Natural Bridges National Monument). The most frequent trends in snow cover were for (1) day of complete snowmelt, which came later for two parks (Bryce Canyon National Park and Golden Spike National Historic Site), and (2) number of snow-free days, which increased for one park (Canyonlands National Park) and decreased for another (Curecanti National Recreation Area). A statistically significant inverse relationship between day of snowmelt and start of growing season occurred for two parks (Arches National Park and Natural Bridges National Monument).

Type
Published Report
Authors
Thoma, David
Date of Issue
2011-06
Publisher
National Park Service, Natural Resource Stewardship and Science
Units
ARCH , BLCA , BRCA , CANY , CARE , CEBR , COLM , CURE , DINO , FOBU , GOSP , HOVE , NABR , NCPN , NRSS , PISP , TICA , ZION
Keywords
Moderate Resolution Imaging Spectrometer, MODIS, Monitoring, NCPN Land Surface Monitoring, NDVI, Normalized Difference Vegetation Index, Remote Sensing, snow-cover extent, vegetation condition, vegetation phenology

Full text (PDF)

Series

See also

Collections

Browse the catalog