Skip to main content
ARS Home » Plains Area » Las Cruces, New Mexico » Range Management Research » Research » Publications at this Location » Publication #402829
Research Project: Science and Technologies for the Sustainable Management of Western Rangeland Systems

Location: Range Management Research

Title: High-resolution image time series to exploit phenological profiles for mapping ecological states in drylands

Author
item MYERS, EMILY - Orise Fellow
item Browning, Dawn
item Burkett, Laura
item Bestelmeyer, Brandon

Submitted to: Ecological Society of America (ESA)
Publication Type: Abstract Only
Publication Acceptance Date: 5/1/2023
Publication Date: 8/11/2023
Citation: Myers, E.R., Browning, D.M., Burkett, L.M., Bestelmeyer, B.T. 2023. High-resolution image time series to exploit phenological profiles for mapping ecological states in drylands. Ecological Society of America (ESA). Abstract.

Interpretive Summary:

Technical Abstract: In arid rangelands, knowledge of soil and climate-determined land potential (land types or ecological sites) and current vegetation state is used to inform site-specific management decisions. Remote sensing methods are commonly used to map land types and states, but accurate classification in arid ecosystems is challenging due to sparse vegetation cover, high spatial heterogeneity, and high inter-annual variability in cover and production. This study aims to evaluate whether remote sensing combined with image-derived phenology metrics can distinguish land types and ecological states in arid rangelands. We evaluate two image sources: near-surface PhenoCams and Harmonized Landsat Sentinel-2 (HLS), collected between 2014 and 2022 at 12 locations at the Jornada Experimental Range in New Mexico, USA. At each location, we extracted 3-day Green Chromatic Coordinate (GCC) time-series data from PhenoCam imagery and Enhanced Vegetation Index (EVI) time-series data from HLS imagery. Phenology metrics (start and end of season, day and value of peak EVI) were extracted from the time-series data for each year and location. Ground data were used to identify land type, ecological state, and dominant plant cover. Data from nearby rain gauges were used to determine ‘wet’ (average to above-average precipitation) and ‘dry’ (below-average precipitation) water years. Current results show differences in image-derived phenology metrics between grass- and shrub-dominated ecological states and between wet and dry years. On average, shrub-dominated locations had an earlier day of year (DOY) start of season (SOS) during wet years (PhenoCam: 101 ± 39, HLS: 96 ± 45) compared to dry years (PhenoCam: 122 ± 34, HLS: 117 ± 36). Grass-dominated locations had a later SOS in wet years (PhenoCam: 148 ± 34, HLS: 123 ± 40) compared to dry years (PhenoCam: 113 ± 50, HLS: 80 ± 22) when grass SOS was more variable. The peak value of EVI was also significantly higher for grass-dominated locations during wet years (0.44 ± 0.07) than for grass-dominated locations during dry years (0.29 ± 0.02) or shrub-dominated locations during wet (0.31 ± 0.07) or dry (0.30 ± 0.04) years. These findings suggest that the use of phenological metrics with remotely sensed imagery has potential to detect ecological state change in arid rangelands to complement field monitoring data, particularly when used in conjunction with precipitation data to identify wet and dry years.