Accurate estimation of surface precipitation with high spatial and temporal resolution is critical for decision making regarding severe weather and water resources management. Polarimetric weather radar is the main operational instrument used for quantitative precipitation estimation (QPE). However, conventional parametric radar QPE algorithms such as the radar reflectivity (Z) and rain rate (R) relations cannot fully represent clouds and precipitation dynamics due to their dependency on local raindrop size distributions and the inherent parameterization errors. This article develops four deep learning (DL) models for polarimetric radar QPE (i.e., RQPENetD1, RQPENetD2, RQPENetV, RQPENetR) using different core building blocks. In particular, multi-dimensional polarimetric radar observations are utilized as input and surface gauge measurements are used as training labels. The feasibility and performance of these DL models are demonstrated and quantified using U.S. Weather Surveillance Radar - 1988 Doppler (WSR-88D) observations near Melbourne, Florida. The experimental results show that the dense blocks-based models (i.e., RQPENetD1 and RQPENetD2) have better performance than residual blocks, RepVGG blocks-based models (i.e., RQPENetR and RQPENetV) and five traditional Z-R relations. RQPENetD1 has the best quantitative performance scores, with a mean absolute error (MAE) of 1.58 mm, root mean squared error (RMSE) of 2.68 mm, normalized standard error (NSE) of 26%, and correlation of 0.92 for hourly rainfall estimates using independent rain gauge data as references. These results suggest that deep learning performs well in mapping the connection between polarimetric radar observations aloft and surface rainfall.
more »
« less
Ground truthed performance of single‐and dual‐polarized radar rain rates at large ranges
Radar accuracy in estimating qualitative precipitation estimation at distances larger than 120 km degrades rapidly because of increased volume coverage and beam height. The performance of the recently upgraded dual‐polarized technology to the NEXRAD network and its capabilities are in need of further examination, as improved rainfall estimates at large distances would allow for significant hydrological modelling improvements. Parameter based methods were applied to radars from St. Louis (KLSX) and Kansas City (KEAX), Missouri, USA, to test the precision and accuracy of both dual‐ and single‐polarized parameter estimations of precipitation at large distances. Hourly aggregated precipitation data from terrestrial‐based tipping buckets provided ground‐truthed reference data. For all KLSX data tested, an R(Z,ZDR) algorithm provided the smallest absolute error (3.7 mm h−1) and root‐mean‐square‐error (45%) values. For most KEAX data, R(ZDR,KDP) and R(KDP) algorithms performed best, with RMSE values of 37%. With approximately 100 h of precipitation data between April and October of 2014, nearly 800 and 400 mm of precipitation were estimated by radar precipitation algorithms but was not observed by terrestrial‐based precipitation gauges for KLSX and KEAX, respectively. Additionally, nearly 30 and 190 mm of measured precipitation observed by gauges were not detected by the radar rainfall estimates from KLSX and KEAX, respectively. Results improve understanding of radar based precipitation estimates from long ranges thereby advancing applications for hydrometeorological modelling and flood forecasting. Copyright © 2016 John Wiley & Sons, Ltd.
more » « less- Award ID(s):
- 1355406
- NSF-PAR ID:
- 10238667
- Publisher / Repository:
- Wiley Blackwell (John Wiley & Sons)
- Date Published:
- Journal Name:
- Hydrological Processes
- Volume:
- 30
- Issue:
- 20
- ISSN:
- 0885-6087
- Format(s):
- Medium: X Size: p. 3692-3703
- Size(s):
- p. 3692-3703
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
The task of nowcasting by deep learning using multivariate, rather than just reflectivity, is limited by poor interpretability. The previous experiment designed MCT (Multivariate Channel Transformer), a deep learning model capable of nowcasting with dual-polarization radar data. Four analytical methods are designed to further explore the contribution of polarization parameters: (i) Case studies of different meteorological processes. (ii) A permutation test ranking the significance of each variable. (iii) Visualization of the feature maps obtained by forward propagation of the input data. (iv) Data downscaling of polarimetric radar data. The results show that the polarization parameters serve as a guide to predict the location and shape of strong reflectivity, as well as the energy retention of strong echoes at 40-50 dBZ. The contributions of Zdr and Kdp are more evident in the prediction results after 30 min, and the importance of Kdp exceeds that of Zdr in case of strong convective weather.more » « less
-
Abstract. Radar dual-wavelength ratio (DWR) measurements from the Stony Brook RadarObservatory Ka-band scanning polarimetric radar (KASPR, 35 GHz), a W-bandprofiling radar (94 GHz), and a next-generation K-band (24 GHz) micro rainradar (MRRPro) were exploited for ice particle identification using triple-frequency approaches. The results indicated that two of the radarfrequencies (K and Ka band) are not sufficiently separated; thus, thetriple-frequency radar approaches had limited success. On the other hand, ajoint analysis of DWR, mean Doppler velocity (MDV), andpolarimetric radar variables indicated potential in identifying ice particletypes and distinguishing among different ice growth processes and even inrevealing additional microphysical details. We investigated all DWR pairs in conjunction with MDV from the KASPRprofiling measurements and differential reflectivity (ZDR) and specificdifferential phase (KDP) from the KASPR quasi-vertical profiles. TheDWR-versus-MDV diagrams coupled with the polarimetric observables exhibiteddistinct separations of particle populations attributed to different rimedegrees and particle growth processes. In fallstreaks, the 35–94 GHz DWRpair increased with the magnitude of MDV corresponding to the scatteringcalculations for aggregates with lower degrees of riming. The DWR valuesfurther increased at lower altitudes while ZDR slightly decreased,indicating further aggregation. Particle populations with higher rimedegrees had a similar increase in DWR but a 1–1.5 m s−1 largermagnitude of MDV and rapid decreases in KDP and ZDR. The analysisalso depicted the early stage of riming where ZDR increased with theMDV magnitude collocated with small increases in DWR. This approach willimprove quantitative estimations of snow amount and microphysical quantitiessuch as rime mass fraction. The study suggests that triple-frequencymeasurements are not always necessary for in-depth ice microphysical studiesand that dual-frequency polarimetric and Doppler measurements cansuccessfully be used to gain insights into ice hydrometeor microphysics.more » « less
-
In this study, a polarimetric radar forward model operator was developed for the Weather Research and Forecasting (WRF) model that was based on a scattering algorithm using the T-matrix methodology. Three microphysics schemes—Thompson, Morrison 2-moment, and Milbrandt-Yau 2-moment—were supported in the operator. This radar forward operator used the microphysics, thermodynamic, and wind fields from WRF model forecasts to compute horizontal reflectivity, radial velocity, and polarimetric variables including differential reflectivity (ZDR) and specific differential phase (KDP) for S-band radar. A case study with severe convective storms was used to examine the accuracy of the radar operator. Output from the radar operator was compared to real radar observations from the Weather Surveillance Radar–1988 Doppler (WSR-88D) radar. The results showed that the radar forward operator generated realistic polarimetric signatures. The distribution of polarimetric variables agreed well with the hydrometer properties produced by different microphysics schemes. Similar to the observed polarimetric signatures, radar operator output showed ZDR and KDP columns from low-to-mid troposphere, reflecting the large amount of rain within strong updrafts. The Thompson scheme produced a better simulation for the hail storm with a ZDR hole to indicate the existence of graupel in the low troposphere.more » « less
-
In this study, a polarimetric radar forward model operator was developed for the Weather Research and Forecasting (WRF) model that was based on a scattering algorithm using the T-matrix methodology. Three microphysics schemes—Thompson, Morrison 2-moment, and Milbrandt-Yau 2-moment—were supported in the operator. This radar forward operator used the microphysics, thermodynamic, and wind fields from WRF model forecasts to compute horizontal reflectivity, radial velocity, and polarimetric variables including differential reflectivity (ZDR) and specific differential phase (KDP) for S-band radar. A case study with severe convective storms was used to examine the accuracy of the radar operator. Output from the radar operator was compared to real radar observations from the Weather Surveillance Radar–1988 Doppler (WSR-88D) radar. The results showed that the radar forward operator generated realistic polarimetric signatures. The distribution of polarimetric variables agreed well with the hydrometer properties produced by different microphysics schemes. Similar to the observed polarimetric signatures, radar operator output showed ZDR and KDP columns from low-to-mid troposphere, reflecting the large amount of rain within strong updrafts. The Thompson scheme produced a better simulation for the hail storm with a ZDR hole to indicate the existence of graupel in the low troposphere.more » « less
