Abstract

Radar and raingage data collected during the International Field Year for the Great Lakes were used to determine the effect of Lake Ontario on precipitation patterns. Objective analysis techniques were used to combine the radar and gage data.

During the warm season the relatively cold lake frequently suppressed afternoon shower activity, particularly when the showers were not associated with large-scale well-organized weather systems. When the showers were scattered, the land portion of the watershed received 402% more rain than the lake compared to 14% more for widespread rain. During the cold season, the lake frequently stimulated precipitation when the 850 mb temperature was more than 7°C colder than the lake.

While the lake influenced the precipitation patterns for about half the days, the total effect on precipitation amounts was small. The lake-effect days were generally those with small-area average amounts. The total warm season rainfall for land areas within 30 km of the lake was 10% more than the lake. For the cold season, the land received 2% less than the lake. There was an orographic component to the precipitation over the far eastern end of the lake and land. Removal of the orographic component tends to reduce the warm season land-to-lake difference while increasing the cold season difference.

Abstract

Quantitative data collected with the WSR-57 radar at Atlantic City from five rainstorms and two snow-storms are compared with precipitation data from 60 recording rain gages within 100 mi of the radar. Hourly rainfall amounts of from 0.01–0.02 inches are detected by the radar in at least 95 per cent of the cases at all radar ranges out to 70 mi. Hourly amounts of from 0.04–0.05 inches are detected in at least 95 per cent of the cases at all ranges out to 100 mi.

The relationship between radar echo intensity and rainfall rate varies from storm to storm. Although the radar appears to have excellent potential for determination of area-average rainfall, reflectivity measurements provide only coarse estimates of point rainfall intensity. The radar estimates of hourly rainfall averages, over a 750 sq mi area within 60 mi of the radar, are within the confidence limits of the average of 10 gage measurements, when a best-fitting radar-rainfall relationship is used for each storm. Use of one grand average relationship for all storms provides estimates of the average areal rainfall whose accuracy corresponds to those of a single rain gage located near the area center.

An analysis of errors made in transferring PPI photographs to digitized arrays and in measuring the echo intensity in steps of 6 db indicates that a reduction in the size of these errors would not substantially improve the accuracy of the radar measurements.

An important unresolved problem concerns the development of techniques for quick determination, under field conditions, of the most accurate reflectivity-rainfall relationship for a particular storm.

A chart based on the average relationship developed in this study is presented for converting echo intensities measured with a WSR-57 to rainfall intensities.

Abstract

Oklahoma thunderstorm data were used to determine how the estimation of area rainfall by radar can be improved by using one or several raingages. The radar data were collected between 1964 and 1968 with the WSR-57 radar at the National Severe Storms Laboratory, Norman, Okla. The rainfall data were obtained from the Agriculture Research Service's dense network of raingages near Chickasha, Okla.

The improvement of area rainfall measurements by combining radar measurements with discrete raingage measurements is demonstrated. It is shown, for example, that the rms error of radar measurements of storm rainfall amount, for a 1000 mi² area, was reduced by 39% after the radar was calibrated with only one rain-gage. At least four uniformly spaced gages are required to measure storm rainfall amounts for the same area as accurately as the radar calibrated with only one gage. The present network of gages over the United States is approximately one gage per 1000 mi².

The ability of radar to measure rainfall variability accurately has been demonstrated; therefore, it is possible to assess objectively whether a particular gage measurement will be useful for adjusting radar rainfall measurements.

With the recent development of an effective system for automatically digitizing and communicating radar data in a form suitable for computer processing, these findings make possible the development of an operational system for measuring rainfall with an accuracy and timeliness never before achieved.

Abstract

Five tornadoes occurred within a 40 min period on 18 May 1984 in eastern Colorado. The evolution of these tornadoes was documented by a single Doppler radar, research aircraft, mesonetwork and chase team. Three of these tornadoes were narrow (≈300 m), rotating dust columns extending from the surface to cloud base more than 5 km from the nearest precipitation. The Doppler-observed parent circulations were <2 km deep and <1 km in diameter. Tornadoes of this type do relatively minor damage and are frequently called gust front tornadoes or gustnadoes. It is believed this is the first Doppler radar documentation of this tornado type. In an operational environment, even at close radar range, it would be difficult to detect the parent circulation associated with these tornadoes. However, by closely monitoring wind shift boundaries and associated localized strong shear regions, preferred tornado areas can be identified.

The other two tornadoes were associated with condensation funnels and occurred near precipitation. The Doppler parent circulations were deeper and wider than the first three tornadoes but were relatively small compared to many of those reported in the literature. All five of these tornadoes occurred along two wind shift lines near the point where the lines intersected. These lines were of synoptic scale origin—a cold front and a trough line.

The low-level echo structure and wind field associated with the parent storm of the two larger tornadoes closely resemble those described in the literature for supercell storms. While the environmental vertical wind shear was less than normally expected for supercell storms, it is believed that the preexisting boundaries created the necessary vorticity and vorticity production mechanisms for tornadogenesis.

Abstract

The data-rich International H₂O Project (IHOP_2002) experiment is used to study convective storm initiation and subsequent evolution for all days of the experiment. Initiation episodes were almost evenly divided between those triggered along surface-based convergence lines and elevated initiation episodes that showed no associated surface convergence. The elevated episodes occurred mostly at night, and the surface-based episodes occurred during the afternoon and evening. Surface-based initiations were mostly associated with synoptic fronts and gust fronts and less so with drylines and bores. Elevated initiations were frequently associated with observable convergent or confluent features in the Rapid Update Cycle (RUC) wind analysis fields between 900 and 600 hPa. The RUC10 3-h forecast of the precipitation initiation episodes were correct 44% of the time, allowing a tolerance of 250 km in space and for the forecast being early by one period. However, the accuracy was closely tied to the scale of the initiation mechanism, being highest for synoptic frontal features and lowest for gust fronts.

Gust fronts were a primary feature influencing the evolution of the initiated storms. Almost one-half of the storm complexes associated with initiation episodes did not produce surface gust fronts. Storm systems that did not produce gust fronts most often lived 2–6 h while those that did frequently lived at least 8 h. The largest and longest-lived storm complexes had well-developed intense gust fronts that influenced the propagation of the storm system. The RUC10 was generally not successful in forecasting the evolution and motion of the larger, more intense storm complexes; presumably this was because it did not produce strong gust fronts.

Implications for forecasting convective storm initiation and evolution are discussed.

Abstract

Radar and satellite data from the Tropical Rainfall Measuring Mission–Large-Scale Biosphere–Atmosphere (TRMM–LBA) project have been examined to determine causes for convective storm initiation in the southwest Amazon region. The locations and times of storm initiation were based on the National Center for Atmospheric Research (NCAR) S-band dual-polarization Doppler radar (S-Pol). Both the radar and the Geostationary Operational Environmental Satellite-8 (GOES-8) visible data were used to identify cold pools produced by convective precipitation. These data along with high-resolution topographic data were used to determine possible convective storm triggering mechanisms. The terrain elevation varied from 100 to 600 m. Tropical forests cover the area with numerous clear-cut areas used for cattle grazing and farming. This paper presents the results from 5 February 1999. A total of 315 storms were initiated within 130 km of the S-Pol radar. This day was classified as a weak monsoon regime where convection developed in response to the diurnal cycle of solar heating. Scattered shallow cumulus during the morning developed into deep convection by early afternoon. Storm initiation began about 1100 LST and peaked around 1500–1600 LST. The causes of storm initiation were classified into four categories. The most common initiation mechanism was caused by forced lifting by a gust front (GF; 36%). Forcing by terrain (>300 m) without any other triggering mechanism accounted for 21% of the initiations and colliding GFs accounted for 16%. For the remaining 27% a triggering mechanism was not identified. Examination of all days during TRMM–LBA showed that this one detailed study day was representative of many days. A conceptual model of storm initiation and evolution is presented. The results of this study should have implications for other locations when synoptic-scale forcing mechanisms are at a minimum. These results should also have implications for very short-period forecasting techniques in any location where terrain, GFs, and colliding boundaries influence storm evolution.

Abstract

The characteristics and causes of a radar artifact called a flare echo are described. The spike or flare-shaped echo typically has reflectivities <20 dBZ, and approaching Doppler velocities. It extends radially 10–20 km downrange of some intense radar storm echoes. Zrnić recently proposed a three-body scattering scenario to explain its occurrence, which consists of scattering by the hydrometeors to the ground, backscattering by the ground to the hydrometeors and scattering by the hydrometeors to the radar. In addition he developed relationships that predict the behavior of the flare reflectivities and velocities.

The data presented here support Zrnić's three-body scattering explanation and relationship, indicating that the flare echo power is dependent on the inverse cube of the distance from the large hydrometeors to the ground. The flue Doppler velocities depend on the radial velocity and fall speed of the hydrometeors responsible for producing the flare. However, it was found that Zrnić's theory did not fully address anomalies observed for scattering paths directly below the large hydrometeors and the contribution of their radial velocities to the flare velocities.

In this paper flare echo data from Colorado and Alabama are compared. The Colorado flares are typically more intense, extensive, and longer lasting and are highly likely to be associated with large (≥ 2 cm) hail and can thus be used as a warning signature. However, this use is not transferrable to Alabama storms where surface hail rarely occurs with flare echoes. In fact, there is evidence that large raindrops may sometimes cause the flare in Alabama.

The flare echo may cause difficulties for unaware researchers using multiple Doppler techniques to synthesize wind fields. It is also a potential problem for forecasters interpreting the data and computer algorithms searching for velocity features such as downbursts and gust fronts. The flare velocities may prove useful for nowcasting microbursts.

Abstract

Assimilation of radar data is one of the key scientific challenges for numerical weather prediction of convective systems. Considerable progress has been made in recent years including retrieval of boundary layer winds from single-Doppler observations, assimilation of radar observations into convective-scale numerical models for explicit thunderstorm prediction, and assimilation of radar estimates of rainfall and wind into mesoscale models. However, the assimilation of radar data for weather prediction remains an important scientific area that demands further investigation. In this paper, the techniques that are currently being used and have demonstrated potential in radar data assimilation are presented. The progress on the research and applications is described and the future directions and challenges are outlined.

Abstract

Thirty-one microburst-producing storms from northeast Colorado were studied using single and multiple Doppler radar for the purpose of identifying radar signatures that indicated the development of a downdraft capable of producing a microburst. Descending reflectivity cores, increasing radial convergence within cloud, rotation and reflectivity notches were found to be microburst precursors, appearing typically 2–6 min prior to initial surface outflow. Descending maximum reflectivity cores coincident with increasing radial convergence within cloud (3–8 km AGL) or near cloud base is believed to be a good indicator of storm downdraft and microburst predictor, especially when coupled with low θ _e air above cloud base and a dry adiabatic lapse rate below cloud base. Three conceptual models have been drawn, based on the 31 events, to summarize the radar signatures of importance in low, moderate, and high-reflectivity microburst-producing storms.

Experience indicates that Doppler radar may aid in providing 0–10 min nowcasts of microbursts. This requires the rapid perusal and assimilation of a large quantity of radar data by the forecaster. To improve upon this effort, a forecaster-computer environment is proposed to allow the forecaster to readily view radar reflectivity and Doppler velocity information in both unprocessed and analyzed form. Use of multi-image radar displays and time-height profiles of quantitative radar estimates of reflectivity and radial shear are suggested to provide an environment where rapid progress can be made in developing techniques to nowcast microbursts.

Abstract

As part of the International Field Year for the Great Lakes, rainfall measurements were made for hurricane Agnes over the Lake Ontario watershed by two C-band radars and one S-band radar. Two dense networks of raingages were located in radar overlap areas. One of the networks contained a raindrop disdrometer for measuring the number and size of drops.

Comparison between radar and gage rainfall measurements show there is considerable time and space variability in the accuracy of the radar rainfall estimates. Because of the high correlation between radars measuring rain over the same area, it is concluded that this observed variability is due to meteorological causes rather than radar measurement errors. The Z-R relationship as measured by the disdrometer remained constant throughout the storm; thus, it is improbable that the variability in the radar accuracies resulted from changes in the Z-R relationship. Some of the larger radar overestimates could be attributed to evaporation below the radar beam. Because of precipitation growth below the beam, incomplete beam filling, and interception of the freezing level, the radar underestimated the rainfall for radar ranges where the beam was greater than 6000 ft above the earth. For areas where the beam height was less than 6000 ft, 90% of the radar estimates were within a factor of 1.3 of the gage measurements.