Quantifying the Sampling Bias in OCO-2 Observation Modes

Thursday, 18 December 2014
Stephanie M Wuerth1, Inez Y Fung1, Annmarie Eldering2 and Junjie Liu2, (1)University of California Berkeley, Berkeley, CA, United States, (2)NASA Jet Propulsion Laboratory, Pasadena, CA, United States
We present results from our 4D analysis CO2 dataset to estimate the sampling bias resulting from OCO-2 observational strategy. Our CO2 fields are produced by simultaneously assimilating 6-hourly weather observations and AIRS-XCO2 observations into the NCAR global carbon-climate model for the year 2003. An ensemble Kalman filter is used for the data assimilation, with an ensemble size of 64, thus producing estimates of the time-evolving standard deviation and mean for the CO2 fields. From these CO2 fields we sample locations where OCO-2 observes CO2 in its 16-day nadir and glint orbit and compare the mean and variability of this subset to that of the full reanalysis. This serves as an estimate of the expected sampling bias of the instrument. We compare the sampling bias for the 16 day nadir / 16 day glint to other options such as 1 day nadir / 1 day glint.