MLS V2.2 IWC is deduced from cloud-induced radiances (Tcir) retrieved at pressures < 215 hPa using the modeled Tcir-IWC relation. The model assumes the particle size distributions (PSDs) parameterized by McFarquhar and Heymsfield (1997), which is a function of IWC and temperature. The histograms on the right contain both clear and cloudy-sky measurements, where clear-sky samples have a Gaussian-like distribution around the zero. Significant cloud measurements are typically defined as those outside the 3-sigma standard deviation of clear-sky variability, which are roughly 3 mg/m3 at 100-121 hPa and 5 mg/m3 at 146-178 hPa. Averaged IWC maps are made on 5-deg x 5-deg latitude-longitude grids and have taken into account both clear and cloudy-sky samples.

The V2.2 IWC has been validated by Wu et al. (JGR Aura Special Section, 2008). The V2.2 IWC represents a spatially-averaged quantity of which the volume size can be approximated with a rectangular box (~4 km high and ~300 km long along track). Roughly speaking, the dimensions of these rectangular boxes represent the vertical and horizontal (along-track) resolutions of the V2.2 IWC measurements, which are given in Table 4. The V2.2 IWC measurements generally have better precision at low pressures and at high latitudes than in other regions. Cloud inhomogeneity and PSD assumption contribute mostly to the uncertainties of the V2.2 IWC retrieval. Most of the inhomogeneity-induced uncertainties can be reduced through averaging (e.g., in monthly maps) due to randomness of the inhomogeneity. However, inhomogeneity-induced scaling errors in the V2.2 IWC retrieval, ranging from -70% to +80%, can not be reduced by averaging. The inhomogeneity-induced uncertainty is larger at higher pressures and for smaller IWC values. Although the validation was focused on the measurements with IWC < 50 mg/m3 where the Tcir-IWC relations are nearly linear, the data with values > 50 mg/m3 are still qualitatively useful.

Table 4. Summary of estimated MLS V2.2 IWC precision, accuracy, and resolution

a)   These are typical 1s precisions of single IWC measurements where the better values are for the extratropics and the poorer values for the tropics. The precision for a particular measurement must be evaluated on a daily basis using the method described in the text.

b)   Estimated from the V2.2 IWC retrieval uncertainties due to cloud inhomogeneity, modeled Tcir-IWC relations, and PSD parameterizations.

c)   H_║, H_┴  and V denote, respectively, the along-track, cross-track and vertical extent of the atmospheric volume sampled by an individual MLS measurement.

d)   This is the range where the stated precision, accuracy and resolution are applied. IWC values above this range, currently giving qualitative information on cloud ice, require further validation for quantitative interpretation.