Advanced Precipitation and Boundary Layer Data Products Derived from ARM Radar Wind Profilers

Accurate measurements of precipitation vertical structure and storm dynamics are needed to help understand the lifecycle of clouds and to help validate simulations produced by cloud resolving models. Radars are good tools to observe different regions and stages of the cloud lifecycle. Cloud radars can observe non-precipitating and lightly precipitating clouds. Yet, these short-wavelength radars cannot observe heavy precipitation because of signal attenuation. In contrast, long-wavelength radar wind profilers (RWPs) can observe light-to-heavy precipitation, but cannot detect non-precipitating clouds. Because of the complementary nature of short- and long-wavelength radars, the next wave of precipitation research is placing radars operating at different frequencies next to each other to simultaneously resolve the full spectrum of cloud regimes. 

In order for RWP observations to be utilized in this multi-frequency radar precipitation research paradigm, it is paramount that the RWP spectrum moments be accurate and calibrated. Except, the Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) RWP data products stored at the ARM Data

ARM data

Continuous field measurements of atmospheric data collected with Atmospheric Radiation Measurement (ARM) instruments at fixed and varying locations worldwide.

Source: Argonne National Laboratory Source

Center are not calculated correctly and are not calibrated. The main purpose of this proposed project is to develop the software code needed to accurately calculate the RWP spectrum moments from the recorded Doppler velocity spectra and then calibrate the RWP reflectivity moments to surface disdrometer observations.

This proposed research will also use spectral processing techniques to develop improved boundary layer RWP horizontal wind profiles with temporal resolutions on the order of 5 minutes. While horizontal wind estimates are not possible within convective storms due to wind motion variability between the radar pulse volume measurement locations, wind estimates are possible just outside of precipitating events. This implies that inflow and outflow storm dynamics, including cold pool winds, can be studied outside of the precipitating regions by processing RWP spectra over short dwell times, on the order of 5 minutes. These studies are not possible with the hourly wind estimates stored at the ARM Data Center.

Observations from six RWPs will be processed. Namely, the four RWPs deployed at the ARM Southern Great Plains (SGP) field site, one RWP deployed during the Green Ocean Amazon 2014/15 (GoAmazon2014/15) field campaign, and one RWP deployed at the ARM Eastern North Atlantic (ENA) field site. This proposed project will deliver the developed software code, PI Data Products, and documentation to the ARM Data Center.

Estimating accurate spectrum moments, calibrated moments, and improved boundary layer horizontal wind profiles will enable future research projects that will advance our understanding of precipitation microphysics within clouds and wind dynamics around precipitating events. For example, future research projects could include estimating cold pool wind speeds, improving turbulence dissipation rates, or determining boundary layer heights based on fuzzy logic or machine learning

machine learning (ML)

Machine Learning (ML) refers to the field and practice of using algorithms that are able to “learn” by extracting patterns from a large body of data.

Source: Center of Excellence Source

techniques.