Temporal scales of tropospheric CO 2 , precipitation, and ecosystem responses in the central Great Plains Ferdouz V. Cochran ⁎, Nathaniel A. Brunsell Department of Geography, University of Kansas, 1475 Jayhawk Blvd., Lawrence, KS 66045-7613, USA abstract article info Article history: Received 5 November 2011 Received in revised form 5 September 2012 Accepted 11 September 2012 Available online xxxx Keywords: Atmospheric infrared sounder Land–atmosphere interactions Wavelets Relative entropy Natural and anthropogenic sources of CO 2 around the globe contribute to mid-tropospheric concentrations, yet it remains unknown how measurements of mid-tropospheric CO 2 relate to regional ecosystem dynamics. NASA At- mospheric Infrared Sounder (AIRS) measurements of CO 2 concentrations in the mid-troposphere from 2002 to 2010 were examined in relation to precipitation and vegetation phenology across the US Great Plains. Wavelet multi-resolution analysis and the information theory metric of relative entropy were applied to assess regional relationships between mid-tropospheric CO 2 , Normalized Difference Vegetation Index (NDVI), and precipitation (PPT). Results show that AIRS observations of mid-tropospheric CO 2 exchange greater amounts of information with regional PPT and NDVI at seasonal, annual, and longer time scales compared to shorter time scales. PPT and NDVI contribute to mid-tropospheric CO 2 at the 18-month time scale, while spatial patterns seen at this time scale for PPT and mid-tropospheric CO 2 are reflective of the influence of PPT on NDVI at the annual scale. Identification of these dominant temporal scales may facilitate utilization of AIRS CO 2 for monitoring regional source/sink dynamics related to climate and land-use/cover change. © 2012 Elsevier Inc. All rights reserved. 1. Introduction Terrestrial ecosystem dynamics influence the Earth's climate sys- tem (Pielke et al., 1998), which is affected by variations in the concen- tration of atmospheric carbon dioxide (CO 2 ). CO 2 is the primary greenhouse gas (GHG) associated with anthropogenic climate change due to the combustion of fossil fuels (Le Quéré et al., 2009). Geo- graphic distribution of anthropogenic sources from fossil fuel emis- sions and land-use change are difficult to assess due to ecosystem heterogeneity and the complex human and natural processes in- volved in carbon release and uptake (Gurney et al., 2009). A recent study by Xiao et al. (2011) estimates that up to 40% of US fossil fuel emissions may be absorbed by natural ecosystems across the US, where net ecosystem exchange (NEE) of CO 2 results in a carbon sink of approximately 0.63 Pg C yr -1 . Interannual variability of NEE is influenced by complex biogeophysical and biogeochemical process- es related to climate change, extreme weather events, and natural and anthropogenic disturbances. Quantifying local to regional sources and sinks of atmospheric CO 2 around the globe continues to be one of the greatest challenges facing climate science today. Reliable methods of monitoring CO 2 concentra- tions at a regional scale are needed for developing policies to regulate CO 2 emissions (Gurney et al., 2009). Different biomes react uniquely across time and space to climate forcings of precipitation and higher levels of atmospheric CO 2 (Stoy et al., 2009). Global and local policies may be inadequate, even detrimental, for ecosystem functioning at re- gional scales. Observations of atmospheric CO 2 concentrations from sat- ellite instruments hold great potential for filling global data gaps, especially in regions lacking surface-based observations. While satellite data may capture advection of CO 2 at the global scale, it is important to know whether such data can also be used to identify sinks and sources within geopolitical boundaries. In December of 2009, mid-troposphere retrievals of CO 2 concentra- tions from the Atmospheric Infrared Sounder (AIRS) instrument, on- board the NASA Earth Observing System Aqua satellite, were made available. Initial analyses of radiance measurements showed promising results, and comparisons of retrievals to transport models, aircraft flasks, and Fourier Transform Infrared Spectrometer measurements indicated an accuracy of 1–2 ppmv (Chahine et al., 2008; Maddy et al., 2008). In contrast to previous assumptions, AIRS observations analyzed by Chahine et al. (2008) have shown that CO 2 is not well mixed in the mid-troposphere. At this level, tropospheric weather and large-scale cir- culation patterns affect CO 2 distribution. In addition, Jiang et al. (2010) have demonstrated that interannual variability of mid-tropospheric CO 2 is related to the El Niño Southern Oscillation (ENSO). Both natural and anthropogenic sources of CO 2 around the globe contribute to mid-tropospheric concentrations. Atmospheric boundary layer (ABL) mechanisms circulate CO 2 released by plant and microbial respiration as well as anthropogenic emissions, and synoptic systems distribute concentrations along atmospheric fronts. The temporal and spatial scales of CO 2 transport between the land surface, the ABL, and the free troposphere vary widely, and there is limited understanding of the processes involved. So far, it is understood that daily evolution Remote Sensing of Environment 127 (2012) 316–328 ⁎ Corresponding author. Tel.: +1 785 864 2021. E-mail address: ferdouzv@ku.edu (F.V. Cochran). 0034-4257/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.rse.2012.09.012 Contents lists available at SciVerse ScienceDirect Remote Sensing of Environment journal homepage: www.elsevier.com/locate/rse