Search

In situ measurements of ice crystal concentrations and sizes made with aircraft instrumentation over the past two decades have often indicated the presence of numerous relatively small (< 50 m diameter) crystals in cirrus clouds. Further, these measurements frequently indicate that small crystals account for a large fraction of the extinction in cirrus clouds. The fact that the instruments used to make these measurements, such as the Forward Scattering Spectrometer Probe (FSSP) and the Cloud Aerosol Spectrometer (CAS), ingest ice crystals into the sample volume through inlets has led to suspicion that the indications of numerous small ]crystals could be artifacts of large ]crystal shattering on the instrument inlets. We present new aircraft measurements in anvil cirrus sampled during the Tropical Composition, Cloud, and Climate Coupling (TC4) campaign with the 2 ] Dimensional Stereo (2D ]S) probe, which detects particles as small as 10 m. The 2D ]S has detector "arms" instead of an inlet tube. Since the 2D ]S probe surfaces are much further from the sample volume than is the case for the instruments with inlets, it is expected that 2D ]S will be less susceptible to shattering artifacts. In addition, particle inter ]arrival times are used to identify and remove shattering artifacts that occur even with the 2D ]S probe. The number of shattering artifacts identified by the 2D ]S interarrival time analysis ranges from a negligible contribution to an order of magnitude or more enhancement in apparent ice concentration over the natural ice concentration, depending on the abundance of large crystals and the natural small ]crystal concentration. The 2D ]S measurements in tropical anvil cirrus suggest that natural small ]crystal concentrations are typically one to two orders of magnitude lower than those inferred from CAS. The strong correlation between the CAS/2D ]S ratio of small ]crystal concentrations and large ]crystal concentration suggests that the discrepancy is likely caused by shattering of large crystals on the CAS inlet. We argue that past measurements with CAS in cirrus with large crystals present may contain errors due to crystal shattering, and past conclusions derived from these measurements may need to be revisited. Further, we present correlations between CAS spurious concentration and 2D ]S large ]crystal mass from spatially uniform anvil cirrus sampling periods as an approximate guide for estimating quantitative impact of large ]crystal shattering on CAS concentrations in previous datasets. We use radiative transfer calculations to demonstrate that in the maritime anvil cirrus sampled during TC4, small crystals indicated by 2D ]S contribute relatively little cloud extinction, radiative forcing, or radiative heating in the anvils, regardless of anvil age or vertical location in the clouds. While 2D ]S ice concentrations in fresh anvil cirrus may often exceed 1 cm.3, and are observed to exceed 10 cm.3 in turrets, they are typically ~0.1 cm.3 and rarely exceed 1 cm.3 (<1.4% of the time) in aged anvil cirrus. We hypothesize that isolated occurrences of higher ice concentrations in aged anvil cirrus may be caused by ice nucleation driven by either small ]scale convection or gravity waves. It appears that the numerous small crystals detrained from convective updrafts do not persist in the anvil cirrus sampled during TC ]4.

In order for clouds to be more accurately represented in global circulation models (GCM), there is need for improved understanding of the properties of ice such as the total water in ice clouds, called ice water content (IWC), ice particle sizes and their shapes. Improved representation of clouds in models will enable GCMs to better predict for example, how changes in emissions of pollutants affect cloud formation and evolution, upper tropospheric water vapor, and the radiative budget of the atmosphere that is crucial for climate change studies. An extensive cloud measurement campaign called CRYSTAL-FACE was conducted during Summer 2002 using instrumented aircraft and a variety of instruments to measure properties of ice clouds. This paper deals with the measurement of IWC using the Harvard water vapor and total water instruments on the NASA WB-57 high-altitude aircraft. The IWC is measured directly by these instruments at the altitude of the WB-57, and it is compared with remote measurements from the Goddard Cloud Radar System (CRS) on the NASA ER-2. CRS measures vertical profiles of radar reflectivity from which IWC can be estimated at the WB-57 altitude. The IWC measurements obtained from the Harvard instruments and CRS were found to be within 20-30% of each other. Part of this difference was attributed to errors associated with comparing two measurements that are not collocated in time an space since both aircraft were not in identical locations. This study provides some credibility to the Harvard and CRS-derived IWC measurements that are in general difficult to validate except through consistency checks using different measurement approaches.

Tropical Storm Chantal during August 2001 was a storm that failed to intensify over the few days prior to making landfall on the Yucatan Peninsula. An observational study of Tropical Storm Chantal is presented using a diverse dataset including remote and in situ measurements from the NASA ER-2 and DC-8 and the NOAA WP-3D N42RF aircraft and satellite. The authors discuss the storm structure from the larger-scale environment down to the convective scale. Large vertical shear (850-200-hPa shear magnitude range 8-15 m/s) plays a very important role in preventing Chantal from intensifying. The storm had a poorly defined vortex that only extended up to 5-6-km altitude, and an adjacent intense convective region that comprised a mesoscale convective system (MCS). The entire low-level circulation center was in the rain-free western side of the storm, about 80 km to the west-southwest of the MCS. The MCS appears to have been primarily the result of intense convergence between large-scale, low-level easterly flow with embedded downdrafts, and the cyclonic vortex flow. The individual cells in the MCS such as cell 2 during the period of the observations were extremely intense, with reflectivity core diameters of 10 km and peak updrafts exceeding 20 m/s. Associated with this MCS were two broad subsidence (warm) regions, both of which had portions over the vortex. The first layer near 700 hPa was directly above the vortex and covered most of it. The second layer near 500 hPa was along the forward and right flanks of cell 2 and undercut the anvil divergence region above. There was not much resemblance of these subsidence layers to typical upper-level warm cores in hurricanes that are necessary to support strong surface winds and a low central pressure. The observations are compared to previous studies of weakly sheared storms and modeling studies of shear effects and intensification. The configuration of the convective updrafts, low-level circulation, and lack of vertical coherence between the upper- and lower-level warming regions likely inhibited intensification of Chantal. This configuration is consistent with modeled vortices in sheared environments, which suggest the strongest convection and rain in the downshear left quadrant of the storm, and subsidence in the upshear right quadrant. The vertical shear profile is, however, different from what was assumed in previous modeling in that the winds are strongest in the lowest levels and the deep tropospheric vertical shear is on the order of 10-12 m/s.

A method of retrieving cloud microphysical properties using combined observations from both cloud radar and lidar is introduced. This retrieval makes use of an improvement to the traditional optimal estimation retrieval method, whereby a series of corrections are applied to the state vector during the search for an iterative solution. This allows faster convergence to a solution and is less processor intensive. The method is first applied to a synthetic cloud t o demonstrate its validity, and it is shown that the retrieval reliably reproduces vertical profiles of ice water content. The retrieval method is then applied to radar and lidar observations from the CRYSTAL-FACE experiment, and vertical profiles of ice crystal diameter, number concentration, and ice water content are retrieved for a cirrus cloud layers observed one day of that experiment. The validity of the relationship between visible extinction coefficient and radar reflectivity was examined. While synthetic tests showed such a functional relationship, the measured data only partially supported such a conclusion. This is due to errors in the forward model (as explained above) as well as errors in the data sets, including possible mismatch between lidar and radar profiles or errors in the optical depth. Empirical relationships between number concentrations and mean particle diameter were also examined. The results indicate that a distinct and robust relationship exists between these retrieved quantities and it is argued that such a relationship is more than an artifact of the retrieval process offering insight into the nature of the microphysical processes taking place in cirrus.

During the NASA Cirrus Regional Study of Tropical Anvils and Cirrus Layers-Florida Area Cirrus Experiment (CRYSTAL-FACE) enhanced mixing ratios of nitric oxide were measured in the anvils of thunderstorms and in clear air downwind of storm systems on flights of a Wl3-57F high-altitude aircraft. Mixing ratios greater than l0 - 20 times background were readily observed over distances of 25-120 km due to lightning activity. In many of the Florida storms deposition of NO occurred up to near the tropopause but major deposition usually occurred 1 - 2 km below the tropopause, or mostly within the visible anvil volume formed prior to storm decay. Observations from two storms of very different anvil size and electrical activity allowed estimates of the total mass of NO, vented to the middle and upper troposphere. Using the cloud-to ground (CG) flash accumulations from the National Lightning Detection Network, climatological intra-cloud (IC) to CG ratios, and assuming that CG and IC flashes were of equivalent efficiency for NO production, the ranges of production per flash for a moderate-sized and a large storm were (0.51 - 1.0) x l0(exp 26) and (2.3 - 3.1) x 10(exp 26) molecules NO/flash, respectively. Using the recently determined average global flash rate of 44 8, a gross extrapolation of these two storms to represent possible global annual production rates yield 1.6 - 3.2 and 7.3 - 9.9 Tg(N)/yr, respectively. If the more usual assumption is made that IC efficiency is l/l0th that of CG activity, the ranges of production for the moderate-sized and large storm were (1.3 - 2.7) x l0(exp 26) and (6.0 - 8.1) x l0(exp 26) molecules NO/CG flash, respectively. The estimates from the large storm may be high because there is indirect evidence that the IC/CG ratio was larger than would be derived from climatology. These two storms and others studied did not have flash rates that scaled as approx. H(sup 5) where H is the cloud top altitude. The observed CG flash accumulations and NO(x) mass production estimate for the month of July over the Florida area were compared with a representative 3D global Chemistry-Transport Model (CTMJ that uses the Price et al. lightning parameterization. For two land grid points representing the Florida peninsula the model compared well with the observations: CG flash rates were low by only a factor of approx. 2. When the model grid points included the coastal regions of Florida the flash accumulations were lower than observed by a factor of 3.4 - 4.6. It is recommended that models using the Price et al. parameterization allow any global coastal grid point to maintain the land rather than the marine flash rate parameterization. The convection in this CTM underestimated the actual cloud top heights over Florida by 1 - 2 km and thus the total lightning flash rates and the altitude range of reactive nitrogen deposition. Broad scale (20 - 120 km) median mixing ratios of NO within anvils over Florida were significantly larger than in storms previously investigated over Colorado and New Mexico.

In order to better understand the characteristics of frozen hydrometeors in hurricane systems, computed brightness temperatures were compared with observations of Hurricane Erin from CAMEX-4 radiometers on the ER-2 aircraft. These wideband observations ranged from 10.7 to 183 plus or minus 10 GHz. In order to initialize the cloud profiles used in the radiative transfer calculations, radar reflectivities from an ER-2 instrument were converted to mass contents at specified height levels. Dropsonde data from the ER-2 and modeled Hurricane Andrew data were used to generate profiles of temperature, pressure, and relative humidity and to provide surface wind speed and ocean temperature. Four different ice (size and density) parameterizations were used to compute brightness temperature values. The comparisons between the calculated and observed brightness temperatures show that in general the Grenfell and Warren particle size distribution does the best job of matching the observations for most of the high frequencies. These high frequencies (greater than or equal to 85 GHz) are particularly sensitive to frozen particles. The Sehkon and Srivastava particle size distribution produced brightness temperatures that were too cold in most cases. On the other hand, the fluffy snow particle parameterizations with and without a melting layer produced brightness temperatures that were much warmer than the observations at these high frequencies. All low frequency calculations track the observations well except in the eye of Hurricane Erin where the surface wind speed and/or near surface temperatures were not properly adjusted to warm the brightness temperature calculations to values closer to the observations. Overall, this work shows that selecting a proper ice particle parameterization can be difficult and additional in situ measurements are needed to define appropriate parameterizations.

Remote sensing observations, such as those from AURA, are necessary to understand the role of cirrus in determining the radiative and humidity budgets of the upper troposphere. Using these measurements quantitatively requires comparisons with in situ measurements that have previously been validated. However, a direct comparison of remote and in situ measurements is difficult due to the requirement that the spatial and temporal overlap be sufficient in order to guarantee that both instruments are measuring the same air parcel. A difficult as this might be for gas phase intercomparisons, cloud inhomogeneities significantly exacerbate the problem for cloud ice water content measurements. The CRYSTAL-FACE mission provided an opportunity to assess how well such intercomparisons can be performed and to establish flight plans that will be necessary for validation of future satellite instruments. During CRYSTAL-FACE, remote and in situ instruments were placed on different aircraft (NASA's ER-2 and WB-59, and the two planes flew in tandem so that the in situ payload flew in the field of view of the remote instruments. We show here that, even with this type of careful flight planning, it is not always possible to guarantee that remote and in situ instruments are viewing the same air parcel. We use ice water data derived from the in situ Harvard Total Water (HV-TW) instrument, and the remote Goddard Cloud Radar System (CRS) and show that agreement between HV-TW and CRS is a strong function of the horizontal separation and the time delay between the aircraft transects. We also use a cloud model to simulate possible trajectories through a cloud and evaluate the use of statistical analysis in determining the agreement between the two instruments. This type of analysis should guide flight planning for future intercomparison efforts, whether for aircraft or satellite-borne instrumentation.

On 26 August 1998 during CAMEX-3, Hurricane Bonnie was overflown by the NASA ER-2 aircraft instrumented with the EDOP X-Band radar and other instrumentation. Bonnie was an asymmetric storm on this day with several prominent features during its brush with the East coast. One of these features was strong intrusion of dry air on the west side of the storm. During the interaction of this westerly shear with the storm, the precipitation band rotates cyclonically from northwest to the south of the storm center. In this paper, we examine the structure of this dry intrusion and its effect on the storm. The EDOP observations are analyzed along with dropsondes, flight-level observations (NASA DC-8 and NOAA P3), and satellite data. The most apparent feature of the EDOP observations are the strong shears in the interface between the western eyewall updraft, and the dry intrusion. Mammatus are formed along this interface reminiscent of convective rear inflows, and suggestive of intense drying of the air through subsidence. This observation as well as other aspects of the analysis are related to recent numerical simulations of Hurricane Bonnie.

NASA conducted the Cirrus Regional Study of Tropical Anvils and Cirrus Layers (CRYSTAL) Florida Area Cirrus Experiment (FACE) during July 2002 for improved understanding of tropical cirrus. One of the goals was to improve the understanding of cirrus generation by convective updrafts. The reasons why some convective storms produce extensive cirrus anvils is only partially related to convective instability and the vertical transport ice mass by updrafts. Convective microphysics must also have an important role on cirrus generation, for example, there are hypotheses that homogeneous nucleation in convective updrafts is a major source of anvil ice particles. In this paper, we report on one intense CRYSTAL- FACE convective case on 16 July 2002 that produced extensive anvil.

NASA's 4th Convection and Moisture Experiment (CAMEX-4) focused on Atlantic hurricanes during the 2001 hurricane season and it involved both NASA and NOAA participation. The NASA ER-2 and DC-8 aircraft were instrumented with unique remote sensing instruments to help increase the overall understanding of hurricanes. This paper is concerned about one of the storms studied, Tropical Storm Chantal, that was a weak storm which failed to intense into a hurricane. One of the practical questions of high importance is why some tropical sto~ins intensify into hurricanes, and others remain weak or die altogether. The magnitude of the difference between the horizontal winds at lower levels and upper altitudes in a tropical storm, i.e., the wind shear, is one important quantity that can affect the intensification of a tropical storm. Strong shear as was present during Tropical Storm Chantal s lifetime and it was detrimental to its intensification. The paper presents an analysis of unique aircraft observations collected from Chantal including an on-board radar, radiometers, dropsondes, and flight level measurements. These measurements have enabled us to examine the internal structure of the winds and thermal structure of Chantal. Most of the previous studies have involved intense hurricanes that overcame the effects of shear and this work has provided new insights into what prevents a weaker storm from intensifying. The storm had extremely intense thunderstorms and rainfall, yet its main circulation was confined to low levels of the atmosphere. Chantal's thermal structure was not configured properly for the storm to intensify. It is most typical that huricanes have a warm core structure where warm temperatures in upper levels of a storm s circulation help intensify surface winds and lower its central pressure. Chantal had two weaker warm layers instead of a well-defined warm core. These layers have been related to the horizontal and vertical winds and precipitation structure and have helped us learn more about why this storm didn't develop.

The Pennsylvania State University-National Center for Atmospheric Research mesoscale model MM5 is used to simulate Hurricane Erin (2001) at high resolution (4-km spacing) from its early development as a tropical depression on 7 September 2001, through a period of rapid intensification into a strong hurricane (8-9 September), and finally into a stage during which it maintains its intensity on 10 September. These three stages of development, intensification, and maintenance in the simulation are in good agreement with the observed evolution of Erin. The simulation shows that during the development and early portions of the intensification stages, intensification is favored because the environmental wind shear is weak and the system moves over a warm tongue of water. As Erin intensifies, the wind'shear gradually increases with the approach of an upper-level trough and strengthening of a low-level high pressure system. By 10 September, the wind shear peaks and begins to decrease, the storm moves over slightly cooler waters, and the intensification ends. Important structural changes occur at this time as the outer precipitation shift from the northeastern and eastern sides to the western side of the eye and precipitation begins to surround the entire eye to initiate the development of a secondary wind maximum and an outer eyewall. The simulation is used to investigate the role of vertical wind shear in the changes of the precipitation structure that took place between 9-10 September by examining the effects of both storm-relative flow changes and changes in the shear-induced tilt. Qualitative agreement is found between the divergence pattern and advection of vorticity by the relative flow with convergence (divergence) generally associated with asymmetric inflow (outflow) in the eyewall region. The shift in the outer precipitation is consistent with a shift in the low-level relative inflow from the northeastern to the northwestern side of the storm. The changes in the relative flow are associated with changes in the winds as the hurricane moves relative to the upper tough and the low-level high pressure system. Examination of the shear-induced tilt of the vortex shows that the change in the tilt direction is greater than that of the shear direction as the tilt shifts from a northerly orientation to northwesterly. Consistent with theory for adiabatic vortices, the maximum low-level convergence and upper-level divergence (and the maximum upward motion) occurs in the direction of tilt. Consequently, both mechanisms may play roles in the changes in the precipitation pattern.

Measurements and analysis of the impact of deep convection on tropopause layer moisture are easily confounded by difficulties making precise observations with sufficient spatial coverage before and after convective events and difficulties distinguishing between changes due to local convection versus large-scale advection. The interactions between cloud microphysics and dynamics in the convective transport of moisture into the tropopause layer also result in a sufficiently complex and poorly characterized system to allow for considerable freedom in theoretical models of stratosphere-troposphere exchange. In this work we perform detailed large-eddy simulations with an explicit cloud microphysics model to study the impact of deep convection on tropopause layer moisture profiles observed over southern Florida during CRYSTALFACE. For four days during the campaign (July 11, 16, 28, and 29) we initialize a 100-km square domain with temperature and moisture profiles measured prior to convection at the PARSL ground site, and initiate convection with a warm bubble that produces an anvil at peak elevations in agreement with lidar and radar observations on that day. Comparing the moisture field after the anvils decay with the initial state, we find that convection predominantly moistens the tropopause layer (as defined by minimum temperature and minimum potential temperature lapse rate), although some drying is also predicted in localized layers. We will also present results of sensitivity tests designed to separate the roles of cloud microphysics and dynamics.

A key focus of CRYSTAL-FACE (Cirrus Regional Study of Tropical Anvils and cirrus Layers - Florida Area Cirrus Experiment) was the generation and subsequent evolution of cirrus outflow from deep convective cloud systems. A preliminary integrated look at the observations of an extended cirrus anvil cloud system observed on 23 July 2002 will be presented, including lidar and millimeter radar observations from NASA's ER-2 and in- situ observations from NASA's WB-57 and University of North Dakota Citation. The observations will be compared to preliminary results of simulations using 1-D and 2-D high-resolution (100 meter) cloud resolving models. The CRMs explicitly account for cirrus microphysical development by resolving the evolving ice crystal size distribution (bin model) in time and space. Both homogeneous and heterogeneous nucleation are allowed in the model. The CRM simulations are driven using the output of regional simulations using MM5 that produces deep convection similar to what was observed. The MM5 model employs a 2 km inner grid (32 layers) over a 360 km domain, nested within a 6 km grid over a 600 km domain. Initial and boundary conditions for the 36- hour MM5 simulation are taken from NCEP Eta model analysis at 32 km resolution. Key issues to be explored are the settling of the observed anvil versus the model simulations, and comparisons of dynamical properties, such as vertical motions, occurring in the observations and models. The former provides an integrated measure of the validity of the model microphysics (fallspeed) while the latter is the key factor in forcing continued ice generation.

A combination multi-aircraft and several satellite sensors were used to examine the core of Hurricane Erin on September 10, 2001, as part of the CAMEX4 program. During the first set of aircraft passes, around 1700 UTC, Erin was still at its maximum intensity with a central pressure of 969 hpa and windspeed of 105 kts (54 m/s). The storm was moving slowly northwestward at 4 m/s, over an increasingly colder sea surface. Three instrumented aircraft, the NOAA P3 with radar, the NASA ER- 2 at 19 km, newly equipped with GPS dropwindsondes, and the NASA DC-8 with dropwindsondes flew in formation across the eye at about 1700 UTC and again 2.5 hrs later around 1930 UTC. The storm had weakened by 13 m/s between the first and second eye penetrations. The warm core had a maximum temperature anomaly of only 11 C, located at 500 hpa, much weaker and lower than active hurricanes. The core appeared to slant rearward above 400 hpa. Even on the first penetration, airborne radar showed that the eye wall cloud towers were dying. The tops fell short of reaching 15 km and a melting band was found throughout. The tropopause had a bulge to 15.8 km elevation (environment approx. 14.4 km) above the dying convection. A feature of Erin at this timt was a pronounced wave-number-one convective asymmetry with all convective activity being confined to the forward quadrants on the left side of the shear vector as calculated from analyses. This is similar to that predicted by the mesoscale numerical models, which also predict that such small amounts of shear would not affect the storm intensity. In Erin, it is remarkable that relatively small shear produced such a pronounced asymmetry in the convection. In addition, horizontal asymmetries in the low-level warm core were identified. Almost certainly, the colder ocean would kill the tall convective towers feeding the warm core, even if wind shear were absent.

The TRMM Precipitation Radar (PR) measures the vertical profile of reflectivity from which the surface rain rate is estimated after attenuation corrections in the 2A21 algorithm. Characteristics of the vertical reflectivity profile is important for various reasons ranging from scientific to instrument algorithms. It is well known that different types of precipitation such as stratiform or convection, have different heating profiles. The vertical profile of reflectivity can provide information on precipitation classification. The vertical reflectivity structure also provides information on precipitation processes such as growth and aggregation. In terms of TRMM algorithms, an independent estimate of the vertical profiles are also extremely important since the PR returns can be attenuated in the rain layer near the surface. Corrections for attenuation are required in the lowest few kilometers, necessitating some assumptions about the rain size distributions and the reflectivity profile below the lowest measurement unaffected by the surface return. Furthermore, some assumptions about the vertical reflectivity profile are required for Ground Validation (GV) radars, since their lowest scan may be 1 or more kilometers above the surface. Statistics on the vertical reflectivity and Doppler structure are presented from the ER-2 Doppler Radar (EDOP) which participated in several TRMM field campaigns (TEFLUN-A, TEFLUN-B, and LBA) and CAMEX-3. The ER-2 aircraft overflew diverse precipitation types during these campaigns. EDOP is an X-band (9.6 GHz) radar for which returns are less attenuated than at the TRMM PR frequency. The EDOP profiles are first corrected for attenuation using the SRT method. The data from all the ER-2 campaigns are then classified by type (convection, stratiform, and other) and then statistics were performed on the vertical reflectivity and Doppler profiles in the form of CFAD's. These CFADs are compared and discussed. The computed CFAD's indicate significant differences as a function of precipitation type and location (hurricane versus non-hurricane, Brazil versus Florida). The implications of these profiles will be discussed.

Observations by the airborne X-band Doppler radar (EDOP) and the NCAR S-band polarimetric (S-POL) radar from two field experiments are used to evaluate the Surface ref'ercnce technique (SRT) for measuring the path integrated attenuation (PIA) and to study attenuation in deep convective storms. The EDOP, flying at an altitude of 20 km, uses a nadir beam and a forward pointing beam. It is found that over land, the surface scattering cross-section is highly variable at nadir incidence but relatively stable at forward incidence. It is concluded that measurement by the forward beam provides a viable technique for measuring PIA using the SRT. Vertical profiles of peak attenuation coefficient are derived in vxo deep convective storms by the dual-wavelength method. Using the measured Doppler velocity, the reflectivities at. the two wavelengths, the differential reflectivity and the estimated attenuation coefficients, it is shown that: supercooled drops and dry ice particles probably co-existed above the melting level in regions of updraft, that water-coated partially melted ice particles probably contributed to high attenuation below the melting level, and that the data are not readil~ explained in terms of a gamma function raindrop size distribution.

Observations by the airborne X-band Doppler radar (EDOP) and the NCAR S-band polarimetric (S-Pol) radar from two field experiments are used to evaluate the surface reference technique (SRT) for measuring the path integrated attenuation (PIA) and to study attenuation in deep convective storms. The EDOP, flying at an altitude of 20 km, uses a nadir beam and a forward pointing beam. It is found that over land, the surface scattering cross-section is highly variable at nadir incidence but relatively stable at forward incidence. It is concluded that measurement by the forward beam provides a viable technique for measuring PIA using the SRT. Vertical profiles of peak attenuation coefficient are derived in two deep convective storms by the dual-wavelength method. Using the measured Doppler velocity, the reflectivities at the two wavelengths, the differential reflectivity and the estimated attenuation coefficients, it is shown that: supercooled drops and (dry) ice particles probably co-existed above the melting level in regions of updraft, that water-coated partially melted ice particles probably contributed to high attenuation below the melting level.

The Precipitation Radar aboard the Tropical Rain Measuring Mission (TRMM) Satellite has shown the potential for spaceborne sensing of snow and rain by means of an incoherent pulsed radar operating at 13.8 GHz. The primary advantage of radar relative to passive instruments arises from the fact that the radar can image the 3-dimensional structure of storms. As a consequence, the radar data can be used to determine the vertical rain structure, rain type (convective/stratiform) effective storm height, and location of the melting layer. The radar, moreover, can be used to detect snow and improve the estimation of rain rate over land. To move toward spaceborne weather radars that can be deployed routinely as part of an instrument set consisting of passive and active sensors will require the development of less expensive, lighter-weight radars that consume less power. At the same time, the addition of a second frequency and an upgrade to Doppler capability are features that are needed to retrieve information on the characteristics of the drop size distribution, vertical air motion and storm dynamics. One approach to the problem is to use a single broad-band transmitter-receiver and antenna where two narrow-band frequencies are spaced apart by 5% to 10% of the center frequency. Use of Ka-band frequencies (26.5 GHz - 40 GHz) affords two advantages: adequate spatial resolution can be attained with a relatively small antenna and the differential reflectivity and mean Doppler signals are directly related to the median mass diameter of the snow and raindrop size distributions. The differential mean Doppler signal has the additional property that this quantity depends only on that part of the radial speed of the hydrometeors that is drop-size dependent. In principle, the mean and differential mean Doppler from a near-nadir viewing radar can be used to retrieve vertical air motion as well as the total mean radial velocity. In the paper, we present theoretical calculations for the differential reflectivity and Doppler as functions of the center frequency, frequency difference, and median mass diameter. For a fixed pair of frequencies, the detectability of the differential signals can be expressed as the number of independent samples required to detect rain or snow with a particular median mass diameter. Because sampling numbers on the order of 1000 are needed to detect the differential signal over a range of size distributions, the instrument must be confined to a near-nadir, narrow swath. Radar measurements from a zenith directed radar operated at 9.1 GHz and 10 GHz are used to investigate the qualitative characteristics of the differential signals. Disdrometer and rain gauge data taken at the surface, just below the radar, are used to test whether the differential signals can be used to estimate characteristics of the raindrop size distribution.

Spaceborne and airborne radars are limited with a respect to the mass and size of the instrument and the power available to operate it. As a consequence, dual-wavelength radars that require separate antennas and power amplifiers are expensive and often impractical. However, if the frequency difference can be reduced so that a single antenna and the same radio-frequency subsystem can be used for both frequencies, dual- wavelength Doppler measurements can be made with a radar of about the same size and mass as its single-frequency counterpart. In the first part of the paper we present calculations of the reflectivity factor differences as functions of the center frequency from 10 to 35 GHz and for frequency differences between -10% and 10% of the center frequency. The results indicate that differential-frequency operation at Ka-band frequencies (26.5 - 40 GHz) provides relatively strong differential signals if the frequencies can be separated by at least 5%. Unlike lower frequency operation, the differential signals at Ka-band (both reflectivity and Doppler) are directly related to the median mass diameter. An important feature of the differential mean Doppler is that it depends only on the drop-size dependent part of the radial velocity. In principle, the mean and mean differential Doppler data from a nadir-looking platform can be used to infer vertical air motion and characteristics of the particle size distribution. To test the instrument concept, the ER-2 Doppler radar was modified for differential frequency operation. Measurements by the modified radar, operating at frequencies of 9.1 GHz and 10 GHz, were made using an 8 degree zenith-pointing offset parabolic antenna. Simultaneous data were taken with an optical rain gauge and an impact disdrometer. Measured and DSD-estimated values of the differential dBZ mean Doppler are presented.

A symposium celebrating the first 50 years of Dr. Joanne Simpson's career took place at the NASA/Goddard Space Flight Center from December 1 - 3, 1999. This symposium consisted of presentations that focused on: historical and personal points of view concerning Dr. Simpson's research career, her interactions with the American Meteorological Society, and her leadership in TRMM; scientific interactions with Dr. Simpson that influenced personal research; research related to observations and modeling of clouds, cloud systems and hurricanes; and research related to the Tropical Rainfall Measuring Mission (TRMM). There were a total of 36 presentations and 103 participants from the US, Japan and Australia. The specific presentations during the symposium are summarized in this paper.

A rich dataset was obtained with observations from the MIR (Millimeter-wave Imaging Radiometer, 89, 150, 183.3$\pm$1, 183.3$\pm$3,183.3$\pm$7, and 220 apprx.GHz), the AMPR (Advanced Microwave Precipitation Radiometer, 10.7, 19.35, 37, and 85 approx. GHz), and the EDOP (ER-2 Doppler Radar, 9.6 approx. GHz) on board the ER-2 aircraft during the CAMEX-3/TEFLUN-B (Convection and Moisture Experiment/Texas and Florida Underflights) TRMM (Tropical Rainfall Measuring Mission) field campaign. Measurements over the ocean from these three instruments on 26 August 1998 were used in our iterative retrieval algorithm to estimate hydrometeor drop size profiles, The algorithm attempts to minimize the difference between the observations and forward radiometer and radar calculations based on the estimated profile. The high frequency MIR observations provide detailed information about the high altitude ice microphysics, while the AMPR is mostly used to define liquid hydrometeor characteristics. The EDOP provides an initial estimate of the profile and as a consistency check throughout the iterative cycle. The retrieval algorithm, specific results for convective and anvil cases, and general implications of this work will be presented.

The NASA ER-2 and DC-8 aircraft collected remote sensing and in situ data sets from Hurricane Bonnie (23, 24, and 26 August 1998) during the Convection And Moisture Experimental-3 (CAMEX-3). Bonnie was an exceptional case where NASA and NOAA had five aircraft sampling both upper levels and lower altitudes. The ER-2 was instrumented with the ER-2 Doppler XBand radar (EDOP) and several radiometers ranging from visible to lower frequency microwaves. EDOP is a fixed dual-beam radar (nadir and forward-looking beams) which allows computation of both vertical and alongtrack horizontal winds. The hurricane secondary circulation is typically difficult to measure at upper levels due to aircraft altitude limitations and sensitivity of the lower altitude airborne radars. EDOP is in principle, well suited to measure these components of the wind. When ER-2 flies across the approximate center of the hurricane circulation, the along-track winds derived from EDOP, are approximately equal to the hurricane radial flow comprising the secondary circulation. Assuming that the hydrometeor fallspeeds can be approximated, the radial and vertical wind components of the secondary circulation can be measured. Since the hydrometeor motions can be estimated with more confidence in the higher altitude ice regions (i.e., graupel and mixed phase are complicated at lower altitudes), the derived radial and vertical winds have higher accuracy at upper levels. On the other hand, the reflectivities are extremely low at higher altitudes, resulting in fewer Doppler velocity estimates.

The Tropical Rain Measuring Mission (TRMM) conducted several intensive field validation campaigns for improved understanding of Tropical precipitation systems. Two of the campaigns (TEFLUN in Florida and Texas, and LBA in Brazil) utilized: the NASA ER-2 high-altitude (20 km) remote sensing aircraft instrumented with the ER-2 Doppler Radar (EDOP), the University of North Dakota Citation microphysics aircraft, and the NCAR S-POL polarization radar. This paper focuses on EDOP-derived fallspeeds and vertical velocities in the rain regions of two stratiform cases (5 September 1998 along the east coast of Florida, and 17 February 1999 in Amazonia in Brazil). These cases were sampled in situ microphysically by the Citation and reported elsewhere in this meeting; the main emphasis of this paper will be on the airborne radar measurements and inferences from them.

During the TRMM-LBA (Tropical Rainfall Measuring Mission - Large-Scale Biosphere-Atmosphere Experiment in Amazonia) field campaign of January - February 1999, EDOP (ER-2 Doppler Radar), AMPR (Advanced Microwave Precipitation Radiometer), and MIR (Millimeter-wave Imaging Radiometer) on board the NASA ER-2 aircraft made a number of flights over the same Amazon area for studies of precipitation signatures. It is generally perceived that AMPR, with measurements at the frequencies of 10.7, 19.35, 37.0, and 85 GHz, is not sensitive to precipitation over land; a possible exception is detection through electromagnetic wave scattering at 85 GHz by frozen hydrometeors aloft above the freezing level. Analysis of the combined data sets from these instruments shows that, in the Amazon highly forested areas where the surface emissivity is high and uniform, direct detection of rain by a radiometer at frequencies less than or equal to 37 GHz is possible. The detection of rain is reflected by a depression in brightness temperature, which amounts to as much as 20 K at 19.35 GHz. Measurements at higher frequencies by the MIR help delineate the regions of scattering signatures above the freezing level. Implications of the combined wideband measurements from AMPR and MIR will be discussed.

Airborne meteorological radars typically operate at attenuating wavelengths. The path integrated attenuation (PIA) can be estimated using the surface reference technique (SRT). In this method, an initial value is determined for the radar cross section of the earth surface in a rain-free area in relatively close proximity to the rain cloud. During subsequent observations of precipitation any decrease 'in the observed surface cross section from the reference value s assumed to be a result of the two-way attenuation along the propagation path. In this paper we present selected instances of high PIA observed over land by an airborne radar. The observations were taken in Brazil and Florida during TRMM (Tropical Rainfall Measurement Mission) field campaigns. We compared these observations with collocated and nearly simultaneous ground-based radar observations by an S-band radar that is not subject to significant attenuation. In this preliminary evaluation, a systematic difference in the attenuation in the two storms is attributed to a difference in the raindrop size distributions; this is supported by observations of ZDR (differential reflectivity).

A persistent, mesoscale region of intense eyewall convection contained within Hurricane Bonnie on 23 August 1998 is examined from multiple observations synthesized from the NASA ER-2 and DC-8 aircraft. The intense convection occurred late in the day as Bonnie was attaining its minimum central pressure and during a stage when the inner core featured a markedly asymmetric structure. The main purpose of this paper is to describe the internal structure of a convective burst and its relationship to the warm core using a synthesis of high-resolution satellite, aircraft radar, and in situ data. An exceptionally vigorous eyewall tower penetrating to nearly 18 km is described. A second intense eyewall tower, adjacent to the eye, is shown to be associated with a mesoscale subsiding current of air that extends horizontally nearly 20 km into the eye interior. The subsidence occupies a substantial depth within the eye and appears to be a much larger scale feature than the convectively-induced, symmetric overturning which commonly occurs on the upper-level flanks of convective towers in other tropical environments.

In this paper, TRMM (Tropical Rainfall Measuring Mission Satellite) Precipitation Radar (PR) products are evaluated by means of simultaneous comparisons with data from the high-altitude ER-2 Doppler Radar (EDOP), as well as ground-based radars. The comparison is aimed primarily at the vertical reflectivity structure, which is of key importance in TRMM rain type classification and latent heating estimation. The radars used in this study have considerably different viewing geometries and resolutions, demanding non-trivial mapping procedures in common earth-relative coordinates. Mapped vertical cross sections and mean profiles of reflectivity from the PR, EDOP, and ground-based radars are compared for six cases. These cases cover a stratiform frontal rainband, convective cells of various sizes and stages, and a hurricane. For precipitating systems that are large relative to the PR footprint size, PR reflectivity profiles compare very well to high-resolution measurements thresholded to the PR minimum reflectivity, and derived variables such as bright band height and rain types are accurate, even at high PR incidence angles. It was found that for, the PR reflectivity of convective cells small relative to the PR footprint is weaker than in reality. Some of these differences can be explained by non-uniform beam filling. For other cases where strong reflectivity gradients occur within a PR footprint, the reflectivity distribution is spread out due to filtering by the PR antenna illumination pattern. In these cases, rain type classification may err and be biased towards the stratiform type, and the average reflectivity tends to be underestimated. The limited sensitivity of the PR implies that the upper regions of precipitation systems remain undetected and that the PR storm top height estimate is unreliable, usually underestimating the actual storm top height. This applies to all cases but the discrepancy is larger for smaller cells where limited sensitivity is compounded by incomplete beam filling. Users of level three TRMM PR products should be aware of this scale dependency.

Current understanding of landfalling tropical cyclones is limited, especially with regard to convective scale processes. On 22 September 1998 Hurricane Georges made landfall on the island of Hispaniola, leaving behind a trail of death and devastation, largely the result of excessive rainfall, not sea level surge or wind. Detailed airborne measurements were taken as part of the Third Convection and Moisture Experiment (CAMEX-3). Of Particular interest are the ER-2 nadir X-band Doppler radar (EDOP) data, which provide a first-time high-resolution view of the precipitation and airflow changes as a hurricane interacts with mountainous terrain. The circulation of hurricane Georges underwent an obvious transition during landfall, evident in the rapid increase in minimum sea-level pressure, the subsidence of the eyewall anvil, and a decrease in average ice concentrations in the eyewall. The eye, as seen in satellite imagery, disappeared, but contrary to current understanding, this was not due to eyewall contraction but rather to convective eruption within the eye. The main convective event within the eye, with upper-level updraft magnitudes near 20 m/s and 89 GHz brightness temperatures below 100 K, occurred when the eye moved over the Cordillera Central, the island's main mountain chain. The location, intensity and evolution of this convection indicate that it was coupled to the surface orography. It is likely that surface rain rates increased during landfall, because of effective droplet collection, both in the convection and in the more widespread stratiform rainfall areas over the island. Evidence for this is the increase in radar reflectivity below the bright band of 1-2 dB/km down to ground-level. Such increase was absent offshore. Such low-level rain enhancement, which cannot be detected in satellite images of upwelling infrared or microwave radiation, must be due to the ascent of boundary-layer air over the topography.

This brief paper illustrates some profiles of radar reflectivity in various precipitation systems observed during the 1998-99 Tropical Rainfall Measuring Mission (TRMM) field campaigns. the profiles were collected by a nadir-viewing radar on the NASA ER-2 flying around 20 km altitude. The paper focuses on hurricane environments.

Reflectivity data collected by the precipitation radar on board the tropical Rainfall Measuring Mission (TRMM) satellite, orbiting at 350 km altitude, are compared to reflectivity data collected nearly simultaneously by a doppler radar aboard the NASA ER-2 flying at 19-20 km altitude, i.e. above even the deepest convection. The TRMM precipitation radar is a scanning device with a ground swath width of 215 km, and has a resolution of about a4.4 km in the horizontal and 250 m in the vertical (125 m in the core swath 48 km wide). The TRMM radar has a wavelength of 217 cm (13.8 GHz) and the Nadir mirror echo below the surface is used to correct reflectivity for loss by attenuation. The ER-2 Doppler radar (EDOP) has two antennas, one pointing to the nadir, 34 degrees forward. The forward pointing beam receives both the normal and the cross-polarized echos, so the linear polarization ratio field can be monitored. EDOP has a wavelength of 3.12 cm (9.6 GHz), a vertical resolution of 37.5 m and a horizontal along-track resolution of about 100 m. The 2-D along track airflow field can be synthesized from the radial velocities of both beams, if a reflectivity-based hydrometer fall speed relation can be assumed. It is primarily the superb vertical resolution that distinguishes EDOP from other ground-based or airborne radars. Two experiments were conducted during 1998 into validate TRMM reflectivity data over convection and convectively-generated stratiform precipitation regions. The Teflun-A (TEXAS-Florida Underflight) experiment, was conducted in April and May and focused on mesoscale convective systems mainly in southeast Texas. TEFLUN-B was conducted in August-September in central Florida, in coordination with CAMEX-3 (Convection and Moisture Experiment). The latter was focused on hurricanes, especially during landfall, whereas TEFLUN-B concentrated on central; Florida convection, which is largely driven and organized by surface heating and ensuing sea breeze circulations. Both TEFLUN-A and B were amply supported by surface data, in particular a dense raingauge network, a polarization radar, wind profilers, a mobile radiosonde system, a cloud physics aircraft penetrating the overflown storms, and a network of 10 cm Doppler radars(WSR-88D). This presentation will show some preliminary comparisons between TRMM, EDOP, and WSR-88D reflectivity fields in the case of an MCS, a hurricane, and less organized convection in central Florida. A validation of TRMM reflectivity is important, because TRMM's primary objective is to estimate the rainfall climatology with 35 degrees of the equator. Rainfall is estimated from the radar reflectivity, as well from TRMM's Microwave Imager, which measures at 10.7, 19.4, 21.3, 37, and 85.5 GHz over a broader swath (78 km). While the experiments lasted about three months the cumulative period of near simultaneous observations of storms by ground-based, airborne and space borne radars is only about an hour long. Therefore the comparison is case-study-based, not climatological. We will highlight fundamental differences in the typical reflectivity profiles in stratiform regions of MCS's, Florida convection and hurricanes and will explain why Z-R relationships based on ground-based radar data for convective systems over land should be different from those for hurricanes. These catastrophically intense rainfall from hurricane Georges in Hispaniola and from Mitch in Honduras highlights the importance of accurate Z-R relationships, It will be shown that a Z-R relationship that uses the entire reflectivity profile (rather than just a 1 level) works much better in a variety of cases, making an adjustment of the constants for different precipitation system categories redundant.