Wavelet analysis of airborne CO2 measurements and related meteorological parameters over heterogeneous landscapes

In this study, we focus on the spatial variations in CO2 and related meteorological parameters quantified in the planetary boundary layer (PBL) from airborne measurements over Illinois, Ohio and Nebraska, USA during the INTEX-NA campaign, July 8th, 2004. The airborne measurements were conducted during morning hours (8:47 a.m) in Nebraska whereas mostly during afternoon hours (1:00 pm and 1:50 pm) in Illinois and Ohio respectively. We perform wavelet analysis using a continuous wavelet transform and wavelet coherence functions for the CO2 data and underlying meteorological variables to interpret the airborne observations. In addition, we also used LANDSAT derived land use/cover information to relate to CO2 variations observed in the PBL. Maximum CO2 mixing ratios were observed over Nebraska and the lowest CO2 mixing ratios over Illinois followed by Ohio. Spectral decomposition of the CO2 data using scalograms revealed lower frequency signals of shorter duration over Illinois compared to Ohio and Nebraska. Further, the high frequency CO2 data for Illinois showed good cyclicity. The high frequency data in Illinois corresponded to low CO2 values of less than 354 ppm, and the time localization of these frequencies closely matched with corn/soybeans mixed agricultural land use suggesting significant CO2 uptake. Results from the wavelet coherence analysis between the CO2 time series and meteorological parameters (potential temperature, relative humidity, water vapor partial pressure, water vapor mixing ratio, wind speed and infrared surface temperature) revealed significant differences in coherences as a function of sampling time. The scale and time dependent wavelet coherence variations observed for CO2 and meteorological data over three different states were attributed to mesoscale variability including variations in the type of vegetation, topography, land-vegetation contrast, cloud cover, and overall landscape heterogeneity.