Atmospheric teleconnection influence on North American land surface phenology

Matthew P Dannenberg,Taehee Hwang,Erika K Wise,Mark Janko,W Kolby Smith

doi:10.1088/1748-9326/aaa85a

Abstract

Short-term forecasts of vegetation activity are currently not well constrained due largely to our lack of understanding of coupled climate-vegetation dynamics mediated by complex interactions between atmospheric teleconnection patterns. Using ecoregion-scale estimates of North American vegetation activity inferred from remote sensing (1982–2015), we examined seasonal and spatial relationships between land surface phenology and the atmospheric components of five teleconnection patterns over the tropical Pacific, north Pacific, and north Atlantic. Using a set of regression experiments, we also tested for interactions among these teleconnection patterns and assessed predictability of vegetation activity solely based on knowledge of atmospheric teleconnection indices. Autumn-to-winter composites of the Southern Oscillation Index (SOI) were strongly correlated with start of growing season timing, especially in the Pacific Northwest. The two leading modes of north Pacific variability (the Pacific-North American, PNA, and West Pacific patterns) were significantly correlated with start of growing season timing across much of southern Canada and the upper Great Lakes. Regression models based on these Pacific teleconnections were skillful predictors of spring phenology across an east-west swath of temperate and boreal North America, between 40°N–60°N. While the North Atlantic Oscillation (NAO) was not strongly correlated with start of growing season timing on its own, we found compelling evidence of widespread NAO-SOI and NAO-PNA interaction effects. These results suggest that knowledge of atmospheric conditions over the Pacific and Atlantic Oceans increases the predictability of North American spring phenology. A more robust consideration of the complexity of the atmospheric circulation system, including interactions across multiple ocean basins, is an important step towards accurate forecasts of vegetation activity.

Highlights

Photosynthetic uptake of CO2 by terrestrial vegetation removes approximately 25% of anthropogenic CO2 emissions from the atmosphere (Ciais et al 2013), and the magnitude and seasonality of this vegetation carbon uptake leave distinct fingerprints on atmospheric CO2 concentrations (Tucker et al 1986, Keeling et al 1996, Graven et al 2013, Forkel et al 2016)
The north Atlantic circulation patterns (NAO and East Atlantic (EA)) were not significantly related to start of growing season (SOS) timing in most ecoregions, and the false discovery tests suggest that these correlations were likely spurious (figures 1(d) and (g))
Southern Oscillation Index (SOI) composites ranging from prior year SON through current year FMA were each significantly correlated (p < 0.05) with SOS timing for roughly 15%–21% of North American ecoregions (figures 1(a) and S6), and the overall correlation field was significant at the 95% level even when controlling for false discovery rate

Summary

Introduction

Photosynthetic uptake of CO2 by terrestrial vegetation removes approximately 25% of anthropogenic CO2 emissions from the atmosphere (Ciais et al 2013), and the magnitude (i.e. primary production) and seasonality (i.e. phenology) of this vegetation carbon uptake leave distinct fingerprints on atmospheric CO2 concentrations (Tucker et al 1986, Keeling et al 1996, Graven et al 2013, Forkel et al 2016). Indices describing atmospheric circulation patterns, the El Niñ o–Southern Oscillation (ENSO) and the North Atlantic Oscillation (NAO), are already important features of many seasonal weather forecasting systems (e.g. Fereday et al 2012, Kalra et al 2013) due to their impacts on regional climates and predictive lead-times of up to several months. These teleconnection patterns are typically treated individually or independently of each other, but in reality they describe components of a complex atmospheric circulation system in which the phase of one teleconnection may modulate the climate effects of another (Wise et al 2015). Some of these circulation patterns, ENSO, may be more variable than in the paleoclimate past (Cobb et al 2013, Li et al 2013), a trend that will likely continue under 21st century warming (Power et al 2013, Cai et al 2014), supporting continued research on the role of teleconnection patterns as drivers of terrestrial ecosystem processes

Methods

Results

Discussion

Conclusion