Reducing representativeness and sampling errors in radio occultation–radiosonde comparisons

[摘要] Radio occultation (RO) and radiosonde (RS) comparisons provide a means of analyzing errors associated with both observational systems. Since RO and RS observations are not taken at the exact same time or location, temporal and spatial sampling errors resulting from atmospheric variability can be significant and inhibit error analysis of the observational systems. In addition, the vertical resolutions of RO and RS profiles vary and vertical representativeness errors may also affect the comparison. In RO–RS comparisons, RO observations are co-located with RS profiles within a fixed time window and distance, i.e. within 3–6 h and circles of radii ranging between 100 and 500 km. In this study, we first show that vertical filtering of RO and RS profiles to a common vertical resolution reduces representativeness errors. We then test two methods of reducing horizontal sampling errors during RO–RS comparisons: restricting co-location pairs to within ellipses oriented along the direction of wind flow rather than circles and applying a spatial–temporal sampling correction based on model data. Using data from 2011 to 2014, we compare RO and RS differences at four GCOS Reference Upper-Air Network (GRUAN) RS stations in different climatic locations, in which co-location pairs were constrained to a large circle ( ∼ 666 km radius), small circle ( ∼ 300 km radius), and ellipse parallel to the wind direction ( ∼ 666 km semi-major axis, ∼ 133 km semi-minor axis). We also apply a spatial–temporal sampling correction using European Centre for Medium-Range Weather Forecasts Interim Reanalysis (ERA-Interim) gridded data. Restricting co-locations to within the ellipse reduces root mean square (RMS) refractivity, temperature, and water vapor pressure differences relative to RMS differences within the large circle and produces differences that are comparable to or less than the RMS differences within circles of similar area. Applying the sampling correction shows the most significant reduction in RMS differences, such that RMS differences are nearly identical to the sampling correction regardless of the geometric constraints. We conclude that implementing the spatial–temporal sampling correction using a reliable model will most effectively reduce sampling errors during RO–RS comparisons; however, if a reliable model is not available, restricting spatial comparisons to within an ellipse parallel to the wind flow will reduce sampling errors caused by horizontal atmospheric variability.

[学科分类] 几何与拓扑