Search

The authors present a method aiming to remove the smoothing effect due to the antenna beamwidth in the reconstruction of the rain reflectivity based on radar measurements. The function to be reconstructed is the spatial distribution Z(x, y, h) of the radar reflectivity Z of the precipitation, which is shown to be related by a Fredholm integral equation to the power received by the radar. This integral equation is transformed into a set of linear equations and subsequently solved through a constrained-linear-inversion technique. The method is then applied to the cases of both stratiform and convective rain, leading to an improved estimation of the reflectivity-peak location and intensity, and of the overall reflectivity profile.

Single Doppler radar analysis has been applied to an occluded front associated with the last stages of cyclonic activity over the Norrkoping area in Central Sweden. Validation was conducted of the analysis methods, such as reflectivity contouring and wind calculation from Doppler radial velocity using velocity azimuth display (VAD). The structure of the cloud system was identified as composed of a stratocumulus cloud layer, horizontally divided into two sectors of high reflectivity (>20 dBZ), oriented parallel to the surface occluded front, each extending around 80 km W-E and 50 km S-N. Embedded in these two precipitating areas were local regions of very high reflectivity (up to 50 dBZ), indicating the presence of scattered heavy precipitation and showers. The case study represents a test of the capabilities of single Doppler radar analysis techniques to be introduced in the next future into the operative forecasting environment in Italy.

Thunderstorms in the vicinity of Johannesburg and Pretoria have been tracked by an S-band radar since 1971. A total of 181 individual cells which occurred during 1974/75 was selected for this study. Of these, the numbers that did and did not produce hail on the ground were 77 and 104, respectively. The mean of the maximum heights of hail-producing cells was found to be 6.3 km AGL for the 40 dBZ contour and 9.1 km for echo tops (23 dBZ). These parameters in rain-producing cells reached only 4.5 and 6.4 km, respectively. The mean temperature in the free atmosphere at echo top height was −37°C for hail-producing cells and −20°C for rain-producing cells. No strict correlation was found for the temperature distribution and for the position of the freezing level versus the occurrence of hail on the ground There is at least a 50% chance for thunderstorm cells exceeding 7 km to produce hail on the ground. Probability graphs relating the occurrence of hail on the ground to heights of the 23 and 40 dBZ ref...

Radar reflectivity profiles of 154 hail cells have been investigated. An average profile (the hail profile) has been derived. A global method has been used since the hail profile is designed to be applied to global hail parameters. Maximum reflectivities are found at h0 − 1 km, where h0 is the height of the 0°C isotherm. The change in reflectivity reaches 3 dB km−1 at h0 − 2 km and for 60 dBZ. This finding agrees with a theoretical profile based on the melting of hailstones. Above h0 the radar reflectivities decrease continuously with a rate of −1 to −2 dB km−1.

Conical graupel is modeled using sphere-cone-oblate spheroidal shapes for the purpose of computing their backscattering properties at a wavelength of 10 cm in terms of the radar polarimetric observables, reflectivity factors, differential reflectivities and circular depolarization ratios. A shape distribution based on in situ measurements is used together with gamma (m=0, 2) size distributions in the computations; both wet and dry graupel are considered. Significant differences in the radar observables between the wet and dry cases are noted and the effects of canting on the radar observables are considered. The implications of these results on the differentiation of hydrometeor phase with radar are discussed.

A 1960 35 GHz radar has been modernized through the use of solid state electronics, Dopplerization and improved data-display capabilities. Radars of this frequency are particularly useful for observing the internal structures of clouds and for detecting low concentrations of ice particles in the atmosphere. The minimum effective radar reflectivity factor of a cloud of water drops that is measurable by this radar at a range of 1 km was estimated to be −36 ± 4 dBZ. Simultaneous airborne and radar measurements showed that the radar reflectivity factors for various water clouds determined from radar measurements were generally in good agreement with those derived from in situ measurements of the drop size spectra. These measurements also showed that the radar can detect clouds in which the diameters of the droplets do not exceed ∼27 μm provided there are sufficient concentrations of 10–15 μm diameter droplets. Clouds containing only 1 L−1 of 100 μm diameter ice crystals (corresponding to a mass conce...

In using a dual-wavelength radar system to detect hail, erroneous positive hail signals can result because of the stronger attenuation of the shorter wavelength radar beam. We present a simple technique to correct for attenuation in dual-wavelength analyses. The technique makes use of an attenuation-reflectivity relationship of the form, A = CZxp, where Zx is the S-band reflectivity, C is a coefficient which is determined on a ray-by-ray basis, and p is the exponent, which is assumed to be a constant. In situations where rays of radar data contain a mixture of rain and hail, the attenuation-correction scheme can erroneously apportion more of the attenuation to hail regions rather than to rain regions. The scheme is modified to account for such situations.

The relationships between the radar reflectivity factor Z and significant physical cloud parameters are studied from a dataset collected with an instrumented aircraft in non- or very weakly precipitating warm clouds. The cloud droplet populations for each flight are statistically described from histograms showing the relative importance and the extent of the precipitating and nonprecipitating component in the cloud field. The limit between precipitating and nonprecipitating warm clouds is found to be about Z = 0.032 mm6 m−3 (− 15 dBZ). No clear correlation appears between Z and liquid water content or median volume diameter in cloud samples that contain droplets larger than about 200 μm in diameter. Some relationships are proposed for smaller droplets in natural clouds. No significant difference is found between dynamically induced artificial clouds and natural ones.

A number of physical factors that influence the relation between measured radar reflectivity and surface rainfall are considered both theoretically and through detailed comparisons of radar and raingauge measurements. These factors include natural differences in raindrop-size distributions, enhancement of radar reflectivity by presence of hailstones or melting snow, diminution of reflectivity by downdrafts, and low-level changes in rainfall rate caused by accretion or evaporation. Results of 374 comparisons in twenty storms, which cover a wide variety of synoptic situations and rainfall patterns, are presented. Magnitudes of the effects of the different factors are estimated, and storm types where they are likely to be significant are pointed out. Also, some ways of compensating for the observed effects are suggested.

One of the Soviet seeding criteria used to detect hail cells at an early stage of development is examined. The radar history of 195 strong convective cells is studied and compared with information of a dense hailpad network (323 pads per 1292 km2). It is found that all hail-bearing cells are detected early in their development. About 50% of the 71 cells which met the seeding criterion never produced hail on the ground. This corresponds to a 30% waste of seeding rockets. The criterion which uses six radar parameters can be replaced by a simpler and slightly more efficient criterion which necessitates only the determination of the height of the 45 dBZ contour H45 in relation to the melting level H0. The new criterion is compared with other criteria and its microphysical implications are discussed.

The reflectivity and polarization characteristics of ice crystals as measured with a 1.8-cm radar are reported. Included are measurements on stratiform ice crystal clouds, showing cancelation ratios of 16 to 20 dB and reflectivities of 7 to 28 dBZ. The degree of preferred orientation was approximately 40 to 50%. Some vertical soundings in snow appear to show changes in the snow crystal form with height. Observations of the rapid changes in polarization characteristics in the upper levels of thunderstorms show some cases for which the depolarizing particles are not the same as those which are responsible for the radar backscatter.

Data from the hailpad network of the National Hail Research Experiment were examined in relation to the equivalent radar reflectivity factors recorded in the lowest level sweeps of the radar beam over the pads during hailstorms in 1972 and 1976. The relationship between hail detected at the ground and reflectivity factor was examined for both areal coverage and on a point-by-point basis for each hailpad. The comparisons show that reflectivity factors of 55 dBZ are often measured when no hail is observed at the ground. Rain alone can give rise to reflectivities of this magnitude. The results of the study show that in northeastern Colorado low-level equivalent radar reflectivity factors alone cannot be used to determine the region of hailfall at the ground, nor are they likely to augment quantitative measurements by a ground network of hail sensors. The results found in northeastern Colorado are compared to results from other geographical regions.

The attenuation of 5 cm radar in rain is investigated theoretically for stratiform and thunderstorm drop size distributions. An iterative attenuation estimation scheme is presented. The effects of attenuation on radar precipitation measurements and the capabilities of the attenuation estimation technique are considered for a variety of hypothetical storm sizes and errors in radar calibration, assumed temperature and assumed drop size distribution. This study indicates that 5 cm radar is an adequate precipitation measuring radar for storms under about 50 dBZ, and that if calibrated correctly and used with the iterative attenuation correction scheme, the 5 cm radar can function moderately well up to about 60 dBZ. Radar calibration accuracy is seen to be a limiting criterion for attenuation correction. The results of this study point out the need for raingages in most situations requiring accurate rainfall measurement.

The presence of melting snow in a radar beam produces a highly enhanced return which can lead to large errors in estimates of areal surface rainfall made with radar. This paper describes an algorithm for use in real time to detect the presence of bright bands in conventional (non-Doppler) weather radar data. The algorithm derives values for the characteristic parameters of height and intensity of the bright band which can then be used to calculate a correction. Detection relies on the fact that the bright band causes a peak in the apparent rainfall rate measured by the radar at a range dependent on its height above the radar and the elevation of the radar beam. Analysis of about 270 hours of data collected over a period of five months shows that the algorithm is reliable. It detects the bright-band effect when it occurs and there is sufficient precipitation present and does not raise any false alarms. When the precipitation is patchy it is able to identify the bright band correctly but not with e...

This paper presents an analysis of the accuracy of rain rate estimates from data observed with a radar that has alternating horizontal and vertical polarization. Theoretical accuracies of rain rates from the reflectivity, the differential reflectivity and the differential propagation phase shift are considered via-a-vis the drop size distribution (DSD) variability, using a computer simulation procedure. First measurements of the differential propagation phase shift have been provided by the National Severe Storms Laboratory's dual-polarized radar, in addition to the reflectivity and the differential reflectivity. An examination of the radar data has revealed factors that could affect the rain rate estimates to a greater extent than the often contended DSD variability in the case of differential reflectivity method. Errors caused by sidelobe contamination significantly affect the differential phase shift data, so that a large spatial scale averaging is required to obtain reasonably accurate rain r...

A 10 cm (*) radar has been used on the steerable 25 m diameter aerial at Chilbolton, U.K., to produce detailed threedimensional maps of raincell distribution and intensity. The present paper examines differences in variation of radar reflectivity with height from the model at present recommended by theCcir.Wind shear can have a considerable influence on the distribution of rain as it falls, and may cause a 45° angle of inclination of a cell. This remains constant as the rain cell travels and is usually offset from the direction of travel of the cell. The information on the variation of radar reflectivity with height and the inclination of the rain cells are each important in the field of radio communications.

The kinetic energy of hailfalls is measured with a 10 cm radar. For comparison of the results a total of 204 hailpad stations was used covering an area of 780 km2, i.e., one pad per 3.8 km2. The experimental setup is such that the largest distance between any hailpad and the radar is less than 30 km. The radar is accurately calibrated with raindrop disdrometers and hailstone spectrometers. The radar data consist of PPI pictures obtained at a constant elevation angle of 5.5° with a time resolution of 1 min and a reflectivity resolution of 5 dBZ. The radar reflectivity values are transformed into kinetic energies by means of empirical Z–Ė relations obtained from measured time-resolved hailstone spectra. The data of six severe hailstorms are studied and the following results are obtained: 1) The correlation coefficient of corresponding hailpad and radar derived kinetic energies at single points of a hailstorm is 0.7, a value which can be increased to 0.8 if four-point averages are taken; 2) the glob...

The effects of beam smoothing and post‐detection integration on radar measurements of reflectivity and meteorological quantities are re‐examined for a simple but general model of space variability of reflectivity. The radar equation is derived for the variable field of reflectivity factor and a set of equations is developed giving the true model values of reflectivity, as well as its first and second derivatives, in terms of the measured quantities. Radar measurements are interpreted as averages over the measurement cell and as point values at the centre of the cell. The biases associated with both interpretations for linear, square‐law and logarithmic detectors are studied. It is shown that these biases depend on the parameter inferred from the radar received power (reflectivity factor, rain rate, etc.), the receiver response and the interpretation given to the measurements (point or averages).

A large sample of radar reflectivity data from essentially a full summer of operation was analyzed to determine the horizontal extents and internal structure of rain areas observed at altitude levels from 2 to 10 km. Results are given on the size distributions of individual cells or patches defined by reflectivity thresholds approximately 4 dBz apart, on the dependence of mean size on altitude and reflectivity threshold, and on the interior structure of the patches as characterized by the number and sizes of higher-threshold patches contained in each echo. In a more detailed analysis of internal structure, the data were restricted to convective echoes in which certain prescribed reflectivities, ranging from about 30 to 50 dBz, were exceeded. It was found that the dependence of mean quantities, such as patch area, on reflectivity and altitude could be approximately described by simple functions, but that the scatter of observations about the mean was usually large.

Previous authors have reported significant correlations between the horizontal extent of convective showers or storms and the volume of rain they produce. This paper employs that idea to develop a simplified method for estimating convective rainfall by considering only the horizontal extent and duration of the precipitation. The present analysis is based on rain gage and radar data from an area in western North Dakota. A synoptic adjustment is applied to the radar rain volume estimates. A quantity called the integrated rainfall coverage can be calculated from either gage or radar data and is found to be well correlated with the rain volumes. The maximum echo area during any one scan in one hour seems to be the hourly radar measurement best correlated with the rain volumes. This limited study suggests that the accuracy of the simplified method approaches that of methods using radar reflectivity data and may have operational value in some special situations.

The paper presents quantitative millimeter wave Doppler radar observations of radar reflectivity and air vertical velocity inside shallow cumuli whose vertical development is restricted by a capping inversion layer. The maximum observed radar reflectivity is −34 to −30 dBZ, which is equivalent to a cloud liquid water content of 0.1 to 0.25 g/m³ depending on cloud droplet size. The main updraft region, which coincides roughly with the maximum reflectivity region, occupies the center of the cloud and extends throughout the radar-visible cloud depth. Downdrafts exhibit more complex structure and are found on the cloud edges.

A remote sensing system is proposed for quantitative operational measurement of the amount of liquid per unit volume in heavy rain. In particular, a 6-GHz radiometer in conjunction with a 10-cm wavelength radar is discussed. An experimental example involving a discrete heavy rain cell with a derived rainrate of 150 mm/h measured by a 6-GHz radiometer/3-GHz radar combination is given. The effect of clear air brightness temperatures, from oxygen and water vapor, on the 6-GHz radiometric rain measurements is analyzed, and an equation relating the 6-GHz brightness temperature from the rain to the radar backscattered power is derived. Design factors important in an operational observing system are also discussed. >

Data taken by aC-band conventional weather radar during 1979, 1980 and 1981 summer campaigns conducted in the western Po Valley, have been analysed in comparison with surface reports of hailfalls, indicating the location, time and estimated losses for each event. The aim of this paper is to determine the signature of thunder-storms on radar data,vs. the presence or absence of hail on the ground. Hail percentages of occurrence greater than 75% of all cases are experienced when reflectivity values of 55 dBZ are detected above 8000m. Detection criteria of this kind are easy to implement in real-time surveillance systems and can be used to elaborate short-time warnings.

A rapidly switched dual-polarization radar technique is used for the first time to obtain two-dimensional spatial distributions of the statistical characteristics of the sizes and concentration of rain drops in rain. These have been obtained using a high-resolution 10-cm wavelength radar, and the data have been used to estimate rainfall rates within small volumes to a much greater accuracy than is available from conventional radar measurements. The technique also gives a clear distinction between ice particles and rain drops.

The radar records of rain observed during 1980 by the weather radar facility of Spino d'Adda in north Italy were used to generate a large data base composed of about 16,000 ground rain cells characterized by a reflectivity factor greater than 34 dBZ. This large sample of rain elements, collected at the low altitude of 1.5 km from the ground, was analyzed in order to supply the statistical information required for rain attenuation modeling. The point rainfall rate distribution calculated against this data base, as well as the cumulative distributions of maximum, mean, and root-mean-square of rain rate referred to an individual rain area (cell), were found to be power law functions of the rain rate. The study of the horizontal extent of cells, based on the entire cell collection, leads to the conclusion that the probability distribution of the linear extent D of a rain cell with a given rate intensity R exceeded over the cell is an exponential function of D, while a power law relates the probability distribution of cells with a given size D to the rain rate exceeded in the cell. The empirically derived formula for calculating the absolute probability of a rain cell with a given intensity and size is proposed and found adequate at least for the observed region and period.

The radar echoes obtained from rain are known to be of two types, one of which shows a band of high echo intensity just below freezing level, the other a column structure. These echoes have been examined using both ground and airborne radar. The observations have been supplemented by measurements of the temperature structure of the atmosphere and by flight observations of the conditions within the precipitation areas. The results suggest that the two types of radar echo correspond to two distinct mechanisms of rain formation. The band of high echo intensity occurs when rain forms by the Bergeron process. Flight observations confirm that it is due to the melting of ice particles as they fall past the 0° isotherm. In addition, other radar bands are observed to form at heights in the atmosphere where the temperature is in the vicinity of − 16° C. The properties of these bands are described and it is concluded that they are due to the spontaneous freezing of relatively large drops in the cloud. The column type echoes occur in convective clouds and their characteristics are found to be entirely different. Even in clouds which extend some thousands of feet above freezing level, ice and snow particles are not involved and the precipitation consists entirely of water drops. Their properties are similar to those of rain from non-freezing clouds and are consistent with the rain having formed by the coalescence process.

It is not possible to deduce the size of raindrops from the average intensity of the radar signal produced by the precipitation region. However, if it is assumed that all the drops are of the same size, a relation between the diameter of the drops and the number per unit volume may be obtained. When measurable relative attenuation between radar waves of different lengths occurs, a different relation between drop size and concentration may be derived therefrom. The combination of these two types of measurement yields a single solution for the size and concentration of the drops. The solution can be considered as an approximation to the size of the largest drops present in the precipitation region, since the smaller drops contribute only slightly to the radar echo. Measurable attenuation of 3-cm waves may be expected to occur under two conditions: (a) heavy rain with large drops, and (b) moderately heavy rain with fairly uniform characteristics over a large area. Measurable attenuation of l-cm wave...

Experiments at wavelength 10 cm have verified the theoretical finding that the power obtained from radiation reflected from rain is proportional to Z, the sum of the sixth powers of diameters of raindrops' contained in a representative volume. No value has been obtained for the constant of proportionality. Correlation of Z with, rainfall intensity R (both at ground level) on logarithmic scales shows that, within a factor of about two, Z α R2. Vertical scans of rain storms by radar indicate that variation of rain content with height is moderate and accountable. It may be possible therefore to determine with useful accuracy the intensity of rainfall at a point quite distant (say 100 km) by the radar echo from that point.

The radar reflectivity and attenuation of rain are computed for various empirical drop-size distributions using more accurate and systematic calculations of the back-scatter and attenuation cross sections from Mie theory. Reflectivity-rainfall (Z-R) relations for 3.2 to 10 cm, incorporating deviations from the Rayleigh theory, are found to be not appreciably different from those calculated from the Rayleigh theory. At shorter wave-lengths, however, there is considerable deviation and a single Z-R relation can be used only over relatively small intervals of rain intensity. There is only slight temperature dependence for the Z-R relations. A modified Marshall-Palmer distribution found by Mueller and Jones in Florida showers, with a smaller frequency of small drops, results in Z being directly proportional to R with values greater thin those of the M-P distribution for rain intensities up to about 10 mm hr−1 and less at higher intensifies. Attenuation is roughly proportional to the rain intensity. A...

The effect of oxotremorine (0.5 mg/kg, ip) on the glycogen content of liver and muscle was studied in rats at intervals of 10, 20, 40, and 160 min. Oxotremorine caused a progressive glycogenolysis for 40 min, and the glycogen content approached normal levels after 160 min in these tissues. Pretreatment with propranolol (5 mg/kg, ip) and phenoxybenzamine (DBZ) (5 mg/kg, ip) could not inhibit the glycogenolytic effect induced by oxotremorine.

Accurate calculation of weighted centroid of dBZ (dBZ stands for decibels of reflectivity of a radar signal reflected off a remote object.) is the basis of studying characteristic tracking algorithms on the vector echoes. The authors mainly discussed the calculation methods of weighted centroid of dBZ on vector echo data. Based on the calculation method on raster echo data, some calculation methods of weighted centroid of dBZ on vector echo data were proposed by considering the weights of features’ area and distance from their location to radar center. And then, taking the base reflectivity products of CINRAD/SA weather radar in Meizhou city of China as data sources, the authors transformed this type of echo data into the vector format. In the end, using the vector echo maps above, the weighted centroid of the same area was calculated by the methods above. Comparing the results with the value of the same area on raster echo maps, the calculated results and the sources of error were analyzed and one conclusion was drawn: it is more accurate to take the features’ area and distance to radar center as weights during the calculation of weighted centroid of dBZ on vector echo maps.

Rain cell size statistics as a function of rain rate have been deduced by employing a radar data base of rain reflectivity data acquired over a three-year period at Wallops Island, VA. These cell statistics have important applications in slant path rain attenuation modeling and remote sensing of the earth's surface from space at frequencies above 10 GHz.

Dual polarization radar data (ZH andZDR) are analysed in 4 convective cells. Quite different relationships betweenZH andZDR are induced by the different hydrometer characteristics in these cells. The radar data, compiled with raingage and hailpad data obtained at the ground, show that in the mixed-phase (hail and rain) type of precipitations the differential reflectivityZDR is about 1.5 dB smaller than in the rain type of precipitations, without reaching zero values as in the hail type. In zones of strong precipitations the differential reflectivityZDR increase as the horizontal reflectivityZH, reaching values > 4 dB which indicate the existence of very big drops.

A major objective in the remote sensing of convective storms by radar is the clear and reliable differentiation between regions of hail and regions of rainfall. This report describes the application of the differential reflectivity (Z(DR)) radar technique to the problem of hail detection. The procedure is based on the markedly different polarization-dependent backscatter characteristics of rain and hail. Field experiments conducted in Colorado during the spring of 1983 provide significant evidence of hail detection with this new radar technique.

This paper investigates the impact of sampling errors on the estimation of the parameters of the relation Z = aRb used in the measurement of rainfall (R) from radar reflectivity (Z). Simulation results show that, at levels of sampling errors likely to be observed in real data sets, of the order 100 gauge and radar observation pairs are required to estimate a with an accuracy better than 15% and b better than 10%. Errors in radar rainfall estimates for rain rates less than 15 mm h−1 are insensitive to the Z-R parameters. High-resolution measurements from a vertically pointing radar (300 to 400 m above ground level, 1-minute averages) are compared with raingauge measurements taken at the radar site during a period of widespread rainfall. The radar provided surprisingly poor estimates of rain rates, with mean standard errors of 80% for rain rates less than 2 mm h−1. The auto-correlations of the residuals in the radar estimates were found to be very low after only 2 minutes, leading to the conclusion that radar estimates of rainfall intensities should be made at a high space/time resolution and then averaged so as to reduce measurement errors arising from sampling problems.

Data from the Tropical Rainfall Measuring Mission (TRMM) precipitation radar (PR) and Kwajalein S-band validation radar (KR) agree well for reflectivity exceeding the sensitivity of the PR threshold ( ;17 dBZ). For echoes above this intensity threshold, the products derived from reflectivity, particularly maps of rainfall rate and convective/stratiform classification, compare well, even though slightly different convective‐stratiform separation techniques and different reflectivity‐rainfall rate ( Z‐R) relations are used for the PR and KR. The KR observations indicate the PR misses only 2.3% of near-surface rainfall but 46% of near-surface rain area ( $0 dBZ) because of its 17-dBZ threshold. The PR senses less than 15% of the echo area observed by the KR above 5-km altitude (i.e., above the 08C level). Thus, the PR highly undersamples weaker echoes associated with stratiform rain near the surface and ice particles aloft but still manages to capture most of the near-surface precipitation accumulation. The temporal sampling of the TRMM PR accurately captures the KR’s overall frequency distribution of reflectivity and its subdivision into convective and stratiform components. However, diurnal and latitudinal variations of precipitation in the vicinity of Kwajalein are not well sampled.

Weather radars are used to measure the electromagnetic radiation backscattered by cloud raindrops. Clouds that backscatter more electromagnetic radiation consist of larger droplets of rain and therefore they produce more rain. The idea is to estimate rain rate by using weather radar as an alternative to rain-gauges measuring rainfall on the ground. In an experiment during two days in June and August 1997 over the Italian-Swiss Alps, data from weather radar and surrounding rain-gauges were collected at the same time. The statistical KNN and the neural SOM classifiers were implemented for the classification task using the radar data as input and the rain-gauge measurements as output. The proposed system managed to identify matching pattern waveforms and the rainfall rate on the ground was estimated based on the radar reflectivities with a satisfactory error rate, outperforming the traditional Z/R relationship. It is anticipated that more data, representing a variety of possible meteorological conditions, will lead to improved results. The results in this work show that an estimation of rain rate based on weather radar measurements treated with statistical and neural classifiers is possible.

The Kwajalein, Marshall Islands, Tropical Rainfall Measuring Mission (TRMM) ground validation radar has provided a multiyear three-dimensional radar dataset at an oceanic site. Extensive rain gauge networks are not feasible over the ocean and, hence, are not available to aid in calibrating the radar or determining a conversion from reflectivity to rain rate. This paper describes methods used to ensure the calibration and allow the computation of quantitative rain maps from the radar data without the aid of rain gauges. Calibration adjustments are made by comparison with the TRMM satelliteborne precipitation radar. The additional steps required to convert the calibrated reflectivity to rain maps are the following: correction for the vertical profile of reflectivity below the lowest elevation angle using climatological convective and stratiform reflectivity profiles; conversion of reflectivity ( Z) to rain rate (R) with a relationship based on disdrometer data collected at Kwajalein, and a gap-filling estimate. The time series of rain maps computed by these procedures include low, best, and high estimates to frame the estimated overall uncertainty in the radar rain estimation. The greatest uncertainty of the rain maps lies in the calibration of the radar (630%). The estimation of the low-altitude vertical profile of reflectivity is also a major uncertainty ( 615%). The Z‐R and data-gap uncertainties are relatively minor (65% or less). These uncertainties help to prioritize the issues that need to be addressed to improve quantitative rainfall mapping over the ocean and provide useful bounds when comparing radar-derived rain estimates with other remotely sensed measures of oceanic rain (such as from satellite passive microwave sensors).

Preliminary wave-tank results indicate that radar scatter from water surfaces is severely affected by rain at low but not at high wind speeds. The effect is governed by both the rain rate and droplet size. A simple experiment to check this phenomenon is described.

IN the past few years quantitative measurements have been made of the radar reflectivity of many storms, particularly severe ones in which hail and even tornadoes have occurred. In particular, Donaldson has published and discussed vertical profiles of reflectivity of storms in the eastern United States1. He finds that the absolute magnitude of the reflectivity increases with the severity of the thunderstorm in the sequence rain, hail, large hail and tornadoes; and that in hail and tornado situations, the maximum reflectivity is found aloft at a height which increases with storm severity.

Statistics of radar reflectivity are presented as a function of height. The data were derived from 3-dimensional maps of radar reflectivity obtained with a 10 cm radar on the 25 m diameter aerial at Chilbolton in southern England. The way in which reflectivity varies with height is of considerable importance to those assessing possible interference between earth-to-satellite links and terrestrial links. The present results show some differences from the CCIR model now used for such planning purposes.

An explanation is given for the peak sensitivity around 9?11 GHz to drop-size distribution and temperature of microwave attenuation due to a rain-filled medium. The sensitivity of radar reflectivity, differential radar reflectivity and differential radar cross-section to drop-size and temperature are also given. It is shown that at frequencies in excess of about 25 GHz the vertical radar cross-section of water drops can be greater than the horizontal.