Flask Versus In Situ Results at Tall Tower Sites
B.M. Walsh1, A.E. Andrews2, J.D. Kofler3, J.C. Williams1, P. Lang2, P. Novelli2, J. Higgs1, M. Heller3, K. Sours3, K. Masarie2, D. Chao3 and D. Neff3
1Science and Technology Corporation, Boulder, CO 80305; 303-497-6960, E-mail: Brooke.Walsh@noaa.gov
2NOAA Earth System Research Laboratory, Boulder, CO 80305
3Cooperative Institute for Research in Environmental Sciences, University of Colorado, Boulder, CO 80309
Seven sites in the NOAA ESRL GMD Tall Tower network are equipped with both in situ systems for measuring CO and CO2 and Programmable Flask Packages (PFP). Typically, daily flask pairs are sampled from the highest in situ level at the tower, and capture a snapshot of the planetary boundary layer during the most well-mixed time of the day. The daily samples are analyzed for fifty or more species of atmospheric trace gases and are used as an independent measure of comparison for CO and CO2. Paired samples provide a measure of repeatability for PFP measurements that includes sampling precision, analysis precision and atmospheric variability. PFP and in situ CO and CO2 data from tower sites are included in the automated web-based Inter-Comparison Project, which is updated daily. Recent comparisons of flask versus in situ data have identified leaks in sample tubing and compressor packages, prompting investigation and timely repair. The flask measurements can also provide insight into pollution plumes and other local influences surrounding the tower. Variations between flask and in situ agreement differ from site to site, and seasonally. Analysis of the flask versus in situ statistics, as well as time series comparisons for each tower will be presented. Constraints in measurements, sampling time, and other problematic factors will also be explored. A brief overview of special projects and future instrument development will be presented.
Figure 1. Current 2011 flask (red) and in situ (blue) measurements from the Boulder Atmospheric Observatory (BAO) tower in Erie, CO. On the second graph, the running mean of the flask minus in situ CO2 differences is indicated in red, while differences for individual samples are shown in green.