Emission: a new technique to correct rainfall estimates for attenuation

Attenuation of the radar return is a serious problem in heavy rainfall events and often leads to major underestimates of rainfall when using C-band radars. We propose a new technique to quantify the attenuation of the radar beam which relies on the principle that all attenuators are emitters; accordingly, any attenuating element can be quantified by the increase in the noise detected by the radar receiver. Effectively the radar is being operated as a radiometer. For most atmospheric conditions a 0.5dB two-way attenuation of the radar beam leads to an increase of about 10K in brightness temperature. The equivalent temperature of the receiver noise is about 1000K, so a change of 10K (1%) should be detected by averaging 10,000 noise samples. This is achieved by measuring the noise over a large number of distant gates where no targets are to be expected. The radar is calibrated for every transmitted pulse by injecting a known noise source for a time equivalent to a few gates. This technique has been rolled out across the UK operational radar network during winter 2010-2011. Our study has shown that any detectable rainfall at the radar site results in a measurable increase in brightness temperature for the two highest elevation scans of the radar (typically at 3 and 4 degrees elevation). For old radomes the two way attenuation can be as high as 3dB for just 2mm/hr at the radar site, but for newer radomes is closer to 1dB. In all cases there is a very large azimuthal variation in attenuation suggesting that the wind is affecting the degree of radome wetting. Such large attenuations have a significant impact on rainfall estimates, and may also affect the efficiency of calibration procedures using rain gauges at short ranges.

Heavy rain causing C-band attenuation is rare in the UK during the winter months, but we have detected increased emission at the lower beams restricted to a few degrees in azimuth which coincide with heavy showers at 20-30km range. During the summer months of 2011 this technique will be further evaluated when there are attenuating storms within the range of the radar. The advantage of this approach is that the attenuation measurement needs no assumption of the microphysics of the target, working equally well for rain and wet hail. Current attenuation correction schemes rely on a ‘gate-by-gate' approach or use polarisation parameters. A knowledge of the total attenuation should provide a powerful constraint for both the gate-by-gate schemes which are notoriously unstable, and for polarisation schemes where there is an uncertainty of a factor of two in the coefficient linking differential phase to attenuation.