Characterizing Error due to Non-Uniform Spatial and Temporal Sampling in the SWOOSH Merged Data Set and Implications for Understanding Long-Term Ozone and WV Variability

The Stratospheric Water Vapor and OzOne Satellite Homogenized (SWOOSH) data set is a vertically-resolved monthly-mean gridded data record that merges data from five limb sounding and solar occultation satellite instruments, including the SAGE instruments, UARS MLS, UARS HALOE, and Aura MLS. SWOOSH includes both individual satellite source records as well as a merged data product.

Currently, SWOOSH uncertainty estimates for the merged timeseries includes terms related to the homogenization adjustment, instrument precision, and a rough estimate of statistical sampling uncertainty based on the number of samples in a bin. To date, no effort has been made to quantify or correct for errors due to non-uniform spatial and temporal sampling of the satellites used by SWOOSH.

Here, we sample a climate model with the sampling patterns of the SWOOSH satellites to quantify errors due to spatial and temporal sampling inhomogeneities. Overall, sampling uncertainties are small during the MLS period, but are substantial in the early record. The climate model-sampled fields are used to develop a new method of correcting for sampling error in SWOOSH. The impact of correcting for sampling errors is discussed in the context of understanding long term variability in stratospheric ozone and WV.