skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Attention:

The NSF Public Access Repository (PAR) system and access will be unavailable from 10:00 PM to 12:00 PM ET on Tuesday, March 25 due to maintenance. We apologize for the inconvenience.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Improving Polarimetric Radar-Based Drop Size Distribution Retrieval and Rain Estimation Using a Deep Neural Network
Abstract Raindrop size distributions (DSD) and rain rate have been estimated from polarimetric radar data using different approaches with the accuracy depending on the errors both in the radar measurements and the estimation methods. Herein, a deep neural network (DNN) technique was utilized to improve the estimation of the DSD and rain rate by mitigating these errors. The performance of this approach was evaluated using measurements from a two-dimensional video disdrometer (2DVD) at the Kessler Atmospheric and Ecological Field Station in Oklahoma as ground truth with the results compared against conventional estimation methods for the period 2006–17. Physical parameters (mass-/volume-weighted diameter and liquid water content), rain rate, and polarimetric radar variables (including radar reflectivity and differential reflectivity) were obtained from the DSD data. Three methods—physics-based inversion, empirical formula, and DNN—were applied to two different temporal domains (instantaneous and rain-event average) with three diverse error assumptions (fitting, measurement, and model errors). The DSD retrievals and rain estimates from 18 cases were evaluated by calculating the bias and root-mean-squared error (RMSE). DNN produced the best performance for most cases, with up to a 5% reduction in RMSE when model errors existed. DSD and rain estimated from a nearby polarimetric radar using the empirical and DNN methods were well correlated with the disdrometer observations; the rain-rate estimate bias of the DNN was significantly reduced (3.3% in DNN vs 50.1% in empirical). These results suggest that DNN has advantages over the physics-based and empirical methods in retrieving rain microphysics from radar observations.  more » « less
Award ID(s):
2136161
PAR ID:
10471141
Author(s) / Creator(s):
 ;  ;  ;  ;  ;  ;  ;  
Publisher / Repository:
American Meteorological Society
Date Published:
Journal Name:
Journal of Hydrometeorology
Volume:
24
Issue:
11
ISSN:
1525-755X
Format(s):
Medium: X Size: p. 2057-2073
Size(s):
p. 2057-2073
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ; (, Atmospheric Measurement Techniques)
    null (Ed.)
    Abstract. The lower-order moments of the drop size distribution (DSD) have generally been considered difficult to retrieve accurately from polarimetric radar data because these data are related to higher-order moments. For example, the 4.6th moment is associated with a specific differential phase and the 6th moment with reflectivity and ratio of high-order moments with differential reflectivity. Thus, conventionally, the emphasis has been to estimate rain rate (3.67th moment) or parameters of the exponential or gamma distribution for the DSD. Many double-moment “bulk” microphysical schemes predict the total number concentration (the 0th moment of the DSD, or M0) and the mixing ratio (or equivalently, the 3rd moment M3). Thus, it is difficult to compare the model outputs directly with polarimetric radar observations or, given the model outputs, forward model the radar observables. This article describes the use of double-moment normalization of DSDs and the resulting stable intrinsic shape that can be fitted by the generalized gamma (G-G) distribution. The two reference moments are M3 and M6, which are shown to be retrievable using the X-band radar reflectivity, differential reflectivity, and specific attenuation (from the iterative correction of measured reflectivity Zh using the total Φdp constraint, i.e., the iterative ZPHI method). Along with the climatological shape parameters of the G-G fit to the scaled/normalized DSDs, the lower-order moments are then retrieved more accurately than possible hitherto. The importance of measuring the complete DSD from 0.1 mm onwards is emphasized using, in our case, an optical array probe with 50 µm resolution collocated with a two-dimensional video disdrometer with about 170 µm resolution. This avoids small drop truncation and hence the accurate calculation of lower-order moments. A case study of a complex multi-cell storm which traversed an instrumented site near the CSU-CHILL radar is described for which the moments were retrieved from radar and compared with directly computed moments from the complete spectrum measurements using the aforementioned two disdrometers. Our detailed validation analysis of the radar-retrieved moments showed relative bias of the moments M0 through M2 was <15 % in magnitude, with Pearson’s correlation coefficient >0.9. Both radar measurement and parameterization errors were estimated rigorously. We show that the temporal variation of the radar-retrieved mass-weighted mean diameter with M0 resulted in coherent “time tracks” that can potentially lead to studies of precipitation evolution that have not been possible so far. 
    more » « less
  2. ; ; (, Journal of Atmospheric and Oceanic Technology)
    Abstract Radar retrievals of drop size distribution (DSD) parameters are developed and evaluated over the mountainous Olympic Peninsula of Washington State. The observations used to develop retrievals were collected during the 2015/16 Olympic Mountain Experiment (OLYMPEX) and included the NASA S-band dual-polarimetric (NPOL) radar and a collection of second-generation Particle Size and Velocity (PARSIVEL2) disdrometers over the windward slopes of the barrier. Nonlinear and random forest regressions are applied to the PARSIVEL2 data to develop retrievals for median volume diameter, liquid water content, and rain rate. Improvement in DSD retrieval accuracy, defined by the mean error of the retrieval relative to PARSIVEL2 observations, was achieved when using the random forest model when compared with nonlinear regression. Evaluation of disdrometer observations and the retrievals from NPOL indicate that the radar retrievals can accurately reproduce observed DSDs in this region, including the common wintertime regime of small but numerous raindrops that is important there. NPOL retrievals during the OLYMPEX period are further evaluated using two-dimensional video disdrometers (2DVD) and vertically pointing Micro Rain Radars. Results indicate that radar retrievals using random forests may be skillful in capturing DSD characteristics in the lowest portions of the atmosphere. 
    more » « less
  3. ; ; (, IEEE Transactions on Geoscience and Remote Sensing)
    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
  4. ; ; (, Remote Sensing)
    The raindrop size distribution (DSD) is vital for applications such as quantitative precipitation estimation, understanding microphysical processes, and validation/improvement of two-moment bulk microphysical schemes. We trace the history of the DSD representation and its linkage to polarimetric radar observables from functional forms (exponential, gamma, and generalized gamma models) and its normalization (un-normalized, single/double-moment scaling normalized). The four-parameter generalized gamma model is a good candidate for the optimal representation of the DSD variability. A radar-based disdrometer was found to describe the five archetypical shapes (from Montreal, Canada) consisting of drizzle, the larger precipitation drops and the ‘S’-shaped curvature that occurs frequently in between the drizzle and the larger-sized precipitation. Similar ‘S’-shaped DSDs were reproduced by combining the disdrometric measurements of small-sized drops from an optical array probe and large-sized drops from 2DVD. A unified theory based on the double-moment scaling normalization is described. The theory assumes the multiple power law among moments and DSDs are scaling normalized by the two characteristic parameters which are expressed as a combination of any two moments. The normalized DSDs are remarkably stable. Thus, the mean underlying shape is fitted to the generalized gamma model from which the ‘optimized’ two shape parameters are obtained. The other moments of the distribution are obtained as the product of power laws of the reference moments M3 and M6 along with the two shape parameters. These reference moments can be from dual-polarimetric measurements: M6 from the attenuation-corrected reflectivity and M3 from attenuation-corrected differential reflectivity and the specific differential propagation phase. Thus, all the moments of the distribution can be calculated, and the microphysical evolution of the DSD can be inferred. This is one of the major findings of this article. 
    more » « less
  5. ; ; ; ; (, 15th European Conference on Antennas and Propagation)
    null (Ed.)
    On 9 September 2019, rain-bands of category-1 Hurricane Dorian passed over a ground instrumentation site in Delmarva peninsula, USA. Drop shapes derived from 2D Video Disdrometer measurements at this site were used to compute the S-band radar cross sections (RCS) for horizontal and vertical polarizations for each drop with equi-volume diameter > 2 mm. These are combined with RCS for the smaller drops assuming equilibrium shapes. Radar reflectivity (Zh ) and differential reflectivity (Zdr ) are calculated for each of the 3 minutes throughout the event which lasted for more than 8 hours. These are compared with simultaneous observations from an S-band polarimetric radar 38 km away. The comparisons highlight the impact of large amplitude drop oscillations on Zdr 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email