An assessment of rain “contamination” in ARM two-channel microwave radiometer measurements Roger Marchand1, Casey Wall1*, Wei Zhao1 and Maria Cadeddu2 1University of Washington, 2Argonne National Laboratory, *Presenting Author

Case 1 Case 3 Motivation! ? !! ✔ ? � ✔ 5000 wet-window flag (open/covered) ! Microwave radiometers (MWRs) are the most commonly used and wet-window flag (open/covered) 1500 3000 accurate instruments ARM has to retrieve cloud liquid water path. open MWR open MWR Unfortunately, MWR data are not easily used in precipitating conditions. 1000 500 There are two reasons for this: 5000 ) )

2 1500 2 1. The measurements are “contaminated” by water on the MWR radome. 3000 covered! covered! MWR MWR 2. Precipitating particles can scatter microwave radiation, yet traditional 1000 500 MWR retrievals neglect scattering. 5000 1500 We designed an experiment that alleviates the “wet radome” problem. 3000 1000 500 5000 The Experiment! 1500 !! 3000 500 ! Two MWRs were operated side by side in a “scanning” or tip-cal mode. (g/m Path Water Liquid 1000 Liquid Water Path (g/m Path Water Liquid One MWR was placed under a cover that kept the radome dry while still 5000 permitting measurements away from zenith (photograph below). The 1500 other MWR was operated normally, with the radome exposed to the sky. 3000 We refer to these as the “covered” and “open” MWRs, respectively. 1000 500 Coincident measurements from the 17.8 18.2 18.6 19 19.4 19.8 covered and open MWRs are 17.5 18 18.5 19 compared to estimate contamination Hours (UTC) Hours (UTC) Case 2 due to a wet radome. ✔ wet-window flag (open/covered) 3000 Note the marked times: ◀ Photograph of the covered MWR. The cover is made of 1000 open MWR • ✔ - open and covered MWRs agree despite wet-window flag, likely styrofoam, which is nearly because the radome on the open MWR is dry during light rain and drizzle ) transparent to microwaves. 2 3000 The open MWR was operated covered! • � - open and covered MWRs disagree, likely because of wet-radome to the right of the covered 1000 MWR contamination during heavy rain MWR, just out of the frame. 3000 • ? - difficult to tell if there is wet-radome contamination because the difference between open and covered MWRs is comparable to magnitude 1000 of spatial variability 150 ◀ Cover Test! 100 3000 50 A bucket of water was poured on the 40 (g/m Path Water Liquid 1000 Summary of Preliminary Findings! MWR cover during 35 a warm, clear day. Wet-radome contamination on uncovered MWRs is difficult to quantify 3000 • 30 This plot shows a because of large spatial variability in liquid water path during precipitating 50 time series of 31.4 1000 conditions. Correcting for it seems unlikely. 45 GHz brightness 40 temperature 3 4 5 6 7 • Current methods for keeping MWRs dry appear to work better than we 60 measured at Hours (UTC) expected. Wet-radome contamination is small during drizzle or light-rain 55 different elevation events, suggesting that data are currently being discarded unnecessarily 50 angles. Note that ▲ Case Studies Time series of liquid water path during three during these periods. 70 Brightness Temperature (K) BrightnessTemperature only measure- rain events. Measurements from different elevation angles are 65 ments at zenith shown (top right corner). Dashed and solid lines shown east- • Using a cover could allow for high-quality measurements from MWRs 60 were contaminated. and west-facing measurements, respectively. The wet-window during precipitating conditions, provided that scattering effects are 0 12 24 36 48 flag indicates that a sensor near the radome is wet. When it is accounted for in the retrieval. This would greatly expand the percentage Time (minutes) triggered, measurements are traditionally discarded. of time that high-quality measurements can be made by MWRs.