Search

The purpose of this study is to investigate ways to combine estimates from the Surface Reference and Hitschfeld-Bordan methods into a hybrid path attenuation estimate and to study its performance using dual-frequency radar data provided by the Dual-Frequency Precipitation Radar (DPR) on board the Global Precipitation Mission (GPM) satellite.

A statistical method to reduce the sidelobe clutter of the Ku-band precipitation radar (KuPR) of the Dual-Frequency Precipitation Radar (DPR) on board the Global Precipitation Measurement (GPM) Core Observatory is described and evaluated using DPR observations. The KuPR sidelobe clutter was much more severe than that of the Precipitation Radar on board the Tropical Rainfall Measuring Mission (TRMM), and it has caused the misidentification of precipitation. The statistical method to reduce sidelobe clutter was constructed by subtracting the estimated sidelobe power, based upon a multiple regression model with explanatory variables of the normalized radar cross section (NRCS) of surface, from the received power of the echo. The saturation of the NRCS at near-nadir angles, resulting from strong surface scattering, was considered in the calculation of the regression coefficients.The method was implemented in the KuPR algorithm and applied to KuPR-observed data. It was found that the received power from sidelobe clutter over the ocean was largely reduced by using the developed method, although some of the received power from the sidelobe clutter still remained. From the statistical results of the evaluations, it was shown that the number of KuPR precipitation events in the clutter region, after the method was applied, was comparable to that in the clutter-free region. This confirms the reasonable performance of the method in removing sidelobe clutter. For further improving the effectiveness of the method, it is necessary to improve the consideration of the NRCS saturation, which will be explored in future work.

Algorithms for estimating precipitation rates from spaceborne radar observations of apparent radar reflectivity depend on attenuation correction procedures. The algorithm suite for the Ku-band precipitation radar aboard the Tropical Rainfall Measuring Mission satellite is one such example. The well-known problem of nonuniform beam filling is a source of error in the estimates, especially in regions where intense deep convection occurs. The error is caused by unresolved horizontal variability in precipitation characteristics such as specific attenuation, rain rate, and effective reflectivity factor. This paper proposes the use of vertical decorrelation for correcting the nonuniform beam filling error developed under the assumption of a perfect vertical correlation. Empirical tests conducted using ground-based radar observations in the current simulation study show that decorrelation effects are evident in tilted convective cells. However, the problem of obtaining reasonable estimates of a governing parameter from the satellite data remains unresolved.

Use of dual-wavelength radar, with properly chosen wavelengths, will significantly lessen the ambiguities in the retrieval of microphysical properties of hydrometeors. In this paper, a dual-wavelength algorithm is described to estimate the characteristic parameters of the snow size distributions. An analysis of the computational results, made at X and Ka bands (T-39 airborne radar) and at S and X bands (CP-2 ground-based radar), indicates that valid estimates of the median volume diameter of snow particles, D(sub 0), should be possible if one of the two wavelengths of the radar operates in the non-Rayleigh scattering region. However, the accuracy may be affected to some extent if the shape factors of the Gamma function used for describing the particle distribution are chosen far from the true values or if cloud water attenuation is significant. To examine the validity and accuracy of the dual-wavelength radar algorithms, the algorithms are applied to the data taken from the Convective and Precipitation-Electrification Experiment (CaPE) in 1991, in which the dual-wavelength airborne radar was coordinated with in situ aircraft particle observations and ground-based radar measurements. Having carefully co-registered the data obtained from the different platforms, the airborne radar-derived size distributions are then compared with the in-situ measurements and ground-based radar. Good agreement is found for these comparisons despite the uncertainties resulting from mismatches of the sample volumes among the different sensors as well as spatial and temporal offsets.

The Global Precipitation Measurement (GPM) program is a mission to measure precipitation from space, and is a similar but much expanded mission of the Tropical Rainfall Measuring Mission. Its scope is not limited to scientific research, but includes practical and operational applications such as weather forecasting and water resource management. To meet the requirements of operational use, the GPM uses multiple low-orbiting satellites to increase the sampling frequency and to create three-hourly global rain maps that will be delivered to the world in quasi-real time. A dual-frequency radar (DPR) will be installed on the primary satellite that plays an important role in the whole mission. The DPR will realize measurement of precipitation with high sensitivity, high precision and high resolutions. This paper describes an outline of the GPM program, its issues and the roles and development of the DPR.

Following the scientific success of the Tropical Rainfall Measuring Mission (TRMM) spearheaded by a group of NASA and NASDA scientists, their external scientific collaborators, and additional investigators within the European Union's TRMM Research Program (EUROTRMM), there has been substantial progress towards the development of a new internationally organized, global scale, and satellite-based precipitation measuring mission. The highlights of this newly developing mission are a greatly expanded scope of measuring capability and a more diversified set of science objectives. The mission is called the Global Precipitation Mission (GPM). Notionally, GPM will be a constellation-type mission involving a fleet of nine satellites. In this fleet, one member is referred to as the "core" spacecraft flown in an approximately 70 degree inclined non-sun-synchronous orbit, somewhat similar to TRMM in that it carries both a multi-channel polarized passive microwave radiometer (PMW) and a radar system, but in this case it will be a dual frequency Ku-Ka band radar system enabling explicit measurements of microphysical DSD properties. The remainder of fleet members are eight orbit-synchronized, sun-synchronous "constellation" spacecraft each carrying some type of multi-channel PMW radiometer, enabling no worse than 3-hour diurnal sampling over the entire globe. In this configuration the "core" spacecraft serves as a high quality reference platform for training and calibrating the PMW rain retrieval algorithms used with the "constellation" radiometers. Within NASA, GPM has advanced to the pre-formulation phase which has enabled the initiation of a set of science and technology studies which will help lead to the final mission design some time in the 2003 period. This presentation first provides an overview of the notional GPM program and mission design, including its organizational and programmatic concepts, scientific agenda, expected instrument package, and basic flight architecture. Following this introduction, we focus specifically on the last topic, that being an analysis which leads to an optimal flight architecture dictated in part by science requirements but constrained by allowable orbital mechanics, instrument scan patterns, and antenna aperture properties. Because the optimal architecture involves an interplay between orbit mechanics and instrument specifications, it is important to recognize that in attempting to serve various scientific themes, the final optimal architecture will represent a compromise concerning dynamic range, spatial resolution, sampling interval, pointing, beam coincidence, and measurement uncertainty. Moreover, cost becomes a major factor in seeking the optimal architecture through the pathways of antenna and instrument scan designs, as well as propulsion requirements associated with the orbit heights of various "constellation" members. Although the results presented at the IGARSS-2001 meeting will likely not be the fully refined flight architecture specifications, they are expected to be nearly complete.

The surface reference technique (SRT) has been studied extensively both theoretically and experimentally over the last decade. It is only with the launch of the Tropical Rain Measuring Mission (TRMM) Precipitation Radar, however, that we can begin to test directly whether the technique provides a reliable means of attenuation correction for spaceborne weather radars. Preliminary results indicate that the method yields results that are qualitatively reasonable when the rain rate is moderate or high and when the surface provides a stable reference value. The structure of the normalized radar cross section of the surface (NRCS), however, is highly complex in the sense that the statistics change with background type, incidence angle, location, and season. As a consequence of this, the reliability of the path attenuation estimate is not proportional simply to the amount of rain attenuation but to the rain attenuation relative to the inherent variability of the reference target. To illustrate the behavior of the SRT, several overpasses of the Hurricane Bonnie are shown. We also show that the proper choice of reference data set (spatial, temporal and global) can be critical to the success of the method.

The need to understand the complementarity of the radar and radiometer is important not only to the Tropical Rain Measuring Mission (TRMM) program but to a growing number of multi-instrumented airborne experiment that combine single or dual-frequency radars with multichannel radiometers. The method of analysis used in this study begins with the derivation of dual-wavelength radar equations for the estimation of a two-parameter drop size distribution (DSD). Defining a "storm model" as the set of parameters that characterize snow density, cloud water, water vapor, and features of the melting layer, then to each storm model there will usually correspond a set of range-profiled drop size distributions that are approximate solutions of the radar equations. To test these solutions, a radiative transfer model is used to compute the brightness temperatures for the radiometric frequencies of interest. A storm model or class of storm models is considered optimum if it provides the best reproduction of the radar and radiometer measurements. Tests of the method are made for stratiform rain using simulated storm models as well as measured airborne data. Preliminary results show that the best correspondence between the measured and estimated radar profiles usually can be obtained by using a moderate snow density (0.1-0.2 g/cu cm), the Maxwell-Garnett mixing formula for partially melted hydrometeors (water matrix with snow inclusions), and low to moderate values of the integrated cloud liquid water (less than 1 kg/sq m). The storm-model parameters that yield the best reproductions of the measured radar reflectivity factors also provide brightness temperatures at 10 GHz that agree well with the measurements. On the other hand, the correspondence between the measured and modeled values usually worsens in going to the higher frequency channels at 19 and 34 GHz. In searching for possible reasons for the discrepancies, It is found that changes in the DSD parameter Mu, the radar constants, or the path-integrated attenuation can affect the high frequency channels significantly. In particular, parameters that cause only modest increases in the median mass diameter of the snow, and which have a minor effect on the radar returns or the low frequency brightness temperature, can produce a strong cooling of the 34 GHz brightness temperature.

Estimation of snow characteristics from airborne radar measurements would complement In-situ measurements. While In-situ data provide more detailed information than radar, they are limited in their space-time sampling. In the absence of significant cloud water contents, dual-wavelength radar data can be used to estimate 2 parameters of a drop size distribution if the snow density is assumed. To estimate, rather than assume, a snow density is difficult, however, and represents a major limitation in the radar retrieval. There are a number of ways that this problem can be investigated: direct comparisons with in-situ measurements, examination of the large scale characteristics of the retrievals and their comparison to cloud model outputs, use of LDR measurements, and comparisons to the theoretical results of Passarelli(1978) and others. In this paper we address the first approach and, in part, the second.

Test data are required to test algorithms for the TRMM Precipitation Radar. These data are needed to test the design of the computer codes under development for the operational phase of the mission, and also to test and evaluate alternative or improved precipitation retrieval algorithms. Over a number of years we have developed and used a 3-dimensional radar model for simulating spaceborne precipitation radars. We have adapted this code to produce data files as close as possible to the TRMM file specifications. In this paper, we will describe the model as it is currently implemented, and show some samples of the synthetic data sets.

This paper describes an outline of the algorithm that estimates the instantaneous profiles of the true radar reflectivity factor and rainfall rate from the radar reflectivity profiles observed by the Precipitation Radar (PR) onboard the TRMM satellite. The major challenge of the algorithm lies in the correction of rain attenuation with the non-uniform beam filling effect. The algorithm was tested with synthetic data and the result is shown.

If the average power at the receiver is substantially reduced by heavy rain, the AGC (automatic gain control) circuit of the rain gauge will try to compensate this reduction by increasing the gain. If this happens, then the pulses created by rain drops are amplified more than they should be and the rainfall rate may be overestimated. If the effective diameter (blocking efficiency) of a particle is 2 mm and if the beam width is 2 cm, each particle will reduce the received power by 10 percent when it crosses the beam. Since the beam is blocked by water drops 75 percent of the total time according to the above calculation, the total received power may be reduced by 7.5 percent. To compensate this reduction to the reference value, the gain of amplifier will be increased by 8.1 percent. This increase of gain will increase all pulse sizes by the same fraction and result in the overestimate of the rainfall rate.

This paper briefly reviews several single-frequency rain profiling methods for an airborne or spaceborne radar. The authors describe the different methods from a unified point of view starting from the basic differential equation. This facilitates the comparisons between the methods and also provides a better understanding of the physical and mathematical basis of the methods. The application of several methods to airborne radar data taken during the Convective and Precipitation/Electrification Experiment is shown. Finally, the authors consider a hybrid method that provides a smooth transition between the Hitschfeld-Bordan method, which performs well at low attenuations, and the surface reference method, for which the relative error decreases with increasing path attenuation.

Air-borne radar measurements of thunderstorms were made as part of the CaPE (Convection and Precipitation/Electrification) experiment in Florida in July 1991. The radar has two channels, X-band (10 GHz) and Ka-band (34.5 GHz), and is capable of measuring cross-polarized returns as well as co-polarized returns. In stratiform rain, the cross-polarized components can be observed only at the bright band region and from the surface reflection. The linear depolarization ratios (LDR's) measured at X-band and Ka-band at the bright band are nearly equal. In convective rain, however, the LDR in Ka-band often exceeds the X-band LDR by several dB, and sometimes by more than 10 dB, reaching LDR values of up to -5 dB over heavy convective rain. For randomly oriented hydrometeors, such high LDR values cannot be explained by single scattering from non-spherical scattering particles alone. Because the LDR by single backscatter depends weakly on the wavelength, the difference between the Ka-band and X-band LDR's suggests that multiple scattering effects prevail in the Ka-band LDR. In order to test this inference, the magnitude of the cross-polarized component created by double scattering was calculated using the parameters of the airborne radar, which for both frequencies has beamwidths of 5.1 degrees and pulse widths of 0.5 microsecond. Uniform rain beyond the range of 3 km is assumed.

The vicinity of KSC, where the primary ground truth site of the Tropical Rainfall Measuring Mission (TRMM) program is located, was the focal point of the Convection and Precipitation/Electrification (CaPE) experiment in Jul. and Aug. 1991. In addition to several specialized radars, local coverage was provided by the C-band (5 cm) radar at Patrick AFB. Point measurements of rain rate were provided by tipping bucket rain gage networks. Besides these ground-based activities, airborne radar measurements with X- and Ka-band nadir-looking radars on board an aircraft were also recorded. A unique combination data set of airborne radar observations with ground-based observations was obtained in the summer convective rain regime of central Florida. We present a comparison of these data intending a preliminary validation. A convective rain event was observed simultaneously by all three instrument types on the evening of 27 Jul. 1991. The high resolution aircraft radar was flown over convective cells with tops exceeding 10 km and observed reflectivities of 40 to 50 dBZ at 4 to 5 km altitude, while the low resolution surface radar observed 35 to 55 dBZ echoes and a rain gage indicated maximum surface rain rates exceeding 100 mm/hr. The height profile of reflectivity measured with the airborne radar show an attenuation of 6.5 dB/km (two way) for X-band, corresponding to a rainfall rate of 95 mm/hr.

The vicinity of KSC, where the primary ground truth site of the Tropical Rainfall Measuring Mission (TRMM) program is located, was the focal point of the Convection and Precipitation/Electrification (CaPE) experiment in July and Aug. 1991. In addition to several specialized radars, local coverage was provided by the C-band (5 cm) radar at Patrick AFB. Point measurements of rain rate were provided by tipping bucket rain gage networks. Besides these ground-based activities, airborne radar measurements with X- and Ka-band nadir-looking radars on board an aircraft were also recorded. A unique combination data set of airborne radar observations with ground-based observations was obtained in the summer convective rain regime of central Florida. We present a comparison of these data intending a preliminary validation. A convective rain event was observed simultaneously by all three instrument types on the evening of 27 July 1991. The high resolution aircraft radar was flown over convective cells with tops exceeding 10 km and observed reflectivities of 40 to 50 dBZ at 4 to 5 km altitude, while the low resolution surface radar observed 35 to 55 dBZ echoes and a rain gage indicated maximum surface rain rates exceeding 100 mm/hr. The height profile of reflectivity measured with the airborne radar show an attenuation of 6.5 dB/km (two way) for X-band, corresponding to a rainfall rate of 95 mm/hr.

Experimental data are presented on dual-wavelength scattering characteristics of subtropical rain observed during the convection and precipitation/electrification (CaPE) experiment. Linear depolarization ratios (LDRs) are calculated at both X-band and Ka-band channels. In stratiform rain, the cross-polarized return signal is generally below the noise level except within the melting layer where the maximum LDR at both frequencies is typically between -12 and -10 dB. In convective rain the range profiles of the LDR in X-band and those in Ka-band do not coincide. The X-band LDR values in convective rain typically remain smaller than -25 dB at high altitudes and take somewhat larger values of -25 to -15 dB at low altitudes. The Ka-band LDR values, in contrast, typically increase with radar range, sometimes reaching a level of -3 dB just before the cross-polarized signal falls below the noise level. Such high values of LDR in Ka-band are most likely attributable to multiple scattering effects.