Assessing Uncertainty and Biases in Total Liquid Water Path Retrieved from W-, Ka-, and G-Band Radar/Radiometer Observations from Space

Radar–radiometers that share the same antenna are of great interest due to their matched resolution and complementarity. Here, a spaceborne radar–radiometer simulator and high-resolution cloud model simulations are used to evaluate the impact of various sources of uncertainty and nonuniform beamfilling (NUBF) effects on the total liquid water path (TLWP) inferred from 238-, 94-, and 35.5-GHz brightness temperature (Tb) or path-integrated attenuation (PIA). Two maritime low cloud regimes are considered: drizzling stratocumulus (Sc) and trade wind shallow cumulus (Cu). Instrument precision, the sensitivity of observations to TLWP, and the natural variability of the relationship between the TLWP and the observable within forward simulations collectively determine the retrieval uncertainty of TLWP. Overall, TLWP retrievals are better constrained in Sc regimes due to their homogeneous appearance. The addition of 238-GHz PIA can be particularly advantageous for the Cu regime. A 50% or smaller retrieval uncertainty can be achieved starting at TLWP of 70 g m^-2, compared to more than 500 g m^-2 by the 94-GHz PIA. In addition, instrument precision of Tb worse than 1 K can potentially escalate the uncertainties of TLWP. NUBF effects result in negative biases in observations and subsequent retrievals, the magnitude of which is commensurate with hydrometeor liquid amount and footprint size. Despite the biases in 238-GHz PIA caused by NUBF being the largest among the three frequencies considered, the associated error propagated into TLWP can be significantly lower than the other frequencies due to its high sensitivity.