Erin L. Bunting | USGS Science

USGS2019

Assessing plant production responses to climate across water-limited regions using Google Earth Engine

(Munson) Climate variability and change acting at broad scales can lead to divergent changes in plant production at local scales. Quantifying how production responds to variation in climate at local scales is essential to understand underlying ecological processes and inform land management decision-making, but has historically been limited in spatiotemporal scale based on the use of discrete ground-based measurements or coarse resolution satellite observations. With the advent of cloud-based computing through Google Earth Engine (GEE), production responses to climate can be evaluated across broad landscapes though time at a resolution useful for ecological and land management applications. Here, GEE was employed to synthesize a multi-platform Landsat time series (1988 – 2014) and evaluate relationships between the soil-adjusted vegetation index (a proxy for plant production) and climate across deserts and plant communities of the southwestern U.S. A “climate pivot point” approach was adopted in GEE to assess the trade-off between production responses to increasing wetness and resistances to drought at 30-m resolution. Consistent with a long-term seasonal climate gradient, production was most related to climate variance during the cool-season in the western deserts, during the warm-season in the eastern deserts, and equally related to both seasons within several desert areas. Communities dominated by grasses and deciduous trees displayed large production responses to an increase in wetness and low resistances to water deficit, while shrublands and evergreen woodlands had variable responses and high drought resistances. Production in plant communities that spanned multiple deserts responded differently to seasonal climate variability in each desert. Defining these plant production sensitivities to climate at 30-m resolution in GEE advances forecasts of how long-term climate trajectories may affect carbon storage, wildlife habitat, and the vulnerability of water-limited ecosystems.

Arizona, California, Colorado, Nevada, New Mexico,↗

USGS2018

Plant production responses to precipitation differ along an elevation gradient and are enhanced under extremes

The sensitivity of plant production to precipitation underlies the functioning of ecosystems. Studies that relate long-term mean annual precipitation and production across multiple sites (spatial relationship) or examine interannual linkages within a site (temporal relationship) can reveal biophysical controls over ecosystem function but have limited ability to infer responses to extreme changes in precipitation that may become more common under climate change. To overcome limitations of using a single approach, we integrated satellite- and ground-based estimates of production with a standardized, multi-site precipitation manipulation experiment across a grassland elevation gradient in the southwestern USA. The responsiveness of production to changes in precipitation followed the order: temporal (0.06–0.13 g m −2 mm −1 ) < spatial (0.21 g m −2 mm −1 ) < experimental relationship (0.25–0.42 g m −2 mm −1 ), suggesting that spatial and temporal relationships determined with satellite- and ground-based estimates cannot be extrapolated to determine the effect of extreme events. A strong production response to differences in mean annual precipitation across sites reinforces a regional control of water availability. Interannual sensitivity to precipitation was strongest at the low elevation grasslands, and the high elevation mixed conifer meadow had a large reduction in production in a drought year. Extreme experimental drought strongly reduced production in low elevation grasslands, but water addition had mixed effects. High elevation meadows were insensitive to both extreme drought and water addition. Our results highlight the importance of accounting for extreme climate regimes and site-level factors when scaling climate change effects up to regional and global scales.

Ecosystems↗