Please direct any questions, comments, or suggestions to Larry Carey at
2. Radar Scanning Philosophy
2a. Case Study Mode
2b. Mesoscale Mode
2c. Rain Gauge Mode
The primary scientific goal of the TRMM (Tropical Rainfall Measurement Mission) Post-Launch Field Campaigns (FC) is to improve our understanding of the underlying physical assumptions and sources of error in the Ground Validation (GV) products and in the cloud resolving models (CRM) used to estimate the latent heating profile from the satellite derived three dimensional precipitation structure. In addition, the TRMM FCs will also provide observational data sets during TRMM satellite overpasses from which the initial TRMM instrument algorithms can be validated and perhaps improved. There are two primary post-launch FCs: 1) the "land" campaign will be over Rondonia, Brazil (TRMM-LBA, Nov 98 - Feb 99) and 2) the "oceanic" campaign over Kwajalein Atoll (KWAJEX, Aug 99 - Sep 99). This document will describe the proposed scanning strategy for TRMM-LBA Brazil.
As depicted in Fig. 1, the FCs must obtain observational data which document the thermodynamic, microphysical, kinematic, and electrical structure of a representative sample of tropical convection and their associated meteorological environments at temporal and spatial scales sufficient to compare to the TRMM satellite, GV, and CRM data. In particular, the FCs should provide information on
b. the 3-D flow field
c. lifecycle characteristics
d. independent rainfall estimates
e. convective vs. stratiform partitioning
f. the thermodynamic structure of the atmosphere.
For more details regarding the scientific objectives of the TRMM-LBA field campaign and associated references, click here.
In order to accomplish these data goals, an observational network centered on two precipitation radars will be fielded in Rondonia, Brazil (Fig. 2). The S-pol (S-band, dual polarimetric) and TOGA (C-band, Doppler) radars are organized in a dual-Doppler configuration characterized by a northwest-to-southeast baseline of approximately 60 km. More information on the S-pol site can be found here. The center of the baseline is located about 30 km to the southwest of Ji-Parana. The resulting dual-Doppler region (defined by a 30 degree beam crossing angle as depicted in Fig. 2) is oriented from northeast-to-southwest and includes both partially deforested pasture land and undisturbed rainforest in the Ji-Parana, Jaru, and Pimenta Bueno Basins. Data from both radars will be collected out to a maximum unambiguous range of 150 km, providing approximately 10^4 km^2 of coverage over Rondonia.
In addition to the scanning radars, a dual-wavelength profiler (S-band and UHF) will be located in the southwestern dual-Doppler lobe just off of the baseline (approximately 35 km to the northwest of S-pol). A network of 17 tipping bucket rain gauges and a 2-D Video Disdrometer will be co-located with the profiler. Additional rain gauges (9) will be placed in a second network located about 90 km to the north of the S-pol radar with 4 gauges spaced at equal distances in between. A network of four radiosondes sites will be centered on the S-pol radar as shown. Cloud-to-ground lightning measurements will be made by a network of four ALDF's (Advanced Lightning Direction Finders) approximately centered on the S-pol radar. In addition, two field change meters will be used to measure the total lightning flash rate (total = in-cloud [IC] + CG). A short range flat plate antenna will be located at the profiler with an approximate range of 35 km and a long range flat plate (150 km range) will be sited at the S-pol radar. This ground based network will be supplemented by microphysical and thermodynamic measurements aboard the University of North Dakota Citation II and remote sensing measurements from the NASA ER-2 aircraft. More details regarding the instrumentation network design can be found here.
This combined set of measurements should provide an unprecedented opportunity for the atmospheric science community to characterize the 4-D microphysical, kinematic, and electrical structure of tropical continental convection over Rondonia. The S-pol and TOGA radars will provide the storm wide microphysical and kinematic information required by TRMM to validate and improve the estimation of latent heating over the tropical continents. The dual- polarimetric measurements (Zh, Vr, Zdr, Kdp, LDR, rhoHV) made by S-pol will allow the estimation of precipitation types and amounts throughout the storm echo as the convection evolves through its lifecycle. The dual-Doppler data from the two radars will provide an estimate of the 3-D mesoscale motions as the storms evolve. These radar estimates of kinematic and precipitation characteristics will be supplemented and validated by additional ground based (rain gauges, disdrometer, profiler) and in-situ measurements by the Citation II aircraft. Observations of the IC and CG lightning flash rate will provide unique information regarding the precipitation structure and updraft strength that will complement these other measurements.
The design philosophy behind the radar scan strategy focuses on the scientific goals of the field campaign as summarized above. In order to achieve these goals, three primary scanning modes have been identified:
2. Mesoscale Mode: mesoscale statistical studies, TRMM overpass comparison, general surveillance, widespread convective echo studies
3. Rain Gauge Mode: detailed rainfall studies with the rain gauge networks, disdrometers, and profiler
In order to fulfill the broader scientific goals of the TRMM Ground Validation (GV) efforts and LBA (Large Scale Biosphere Atmosphere Experiment) hydrometeorology program, a low-level 360 degree volume scan will be accomplished every 15 minutes during the Field Campaign by both the S-pol and TOGA radars as depicted in Fig. 3. This low-level volume scan will take 3.5 minutes to complete and will also serve as a surveillance scan during focused sector scanning of an individual cell or groups of convective cells.
The spatial resolution and vertical coverage of the S-pol rain-surveillance scan (RainSurS) can be found in Fig. 4 below. The first two elevation angles are less than 1 deg in order to capture near-surface rainfall. The elevation angle stepping is increased for the final two sweeps in order to increase the sampled height at close-in ranges for surveillance purposes and to compensate for the potential effects of partial beam blocking.
Similarly, a depiction of the TOGA rain-surveillance scan (RainSurT) is given in Fig. 5. Note that RainSurT scan accomplishes 11 low-level sweeps in 3.5 minutes versus only 4 sweeps for the RainSurS scan. The increased spatial sampling of the TOGA radar scan results from the TOGA antenna being rotated at 21 deg/sec during full 360 degree scans while the S-pol antenna rotates at only 7 deg/sec. The S-pol antenna must be rotated slower in order to obtain 64 H/V sample pairs (128 hits) at a PRF of 960 Hz while maintaining an azimuthal spacing of about 0.9 degrees. Esssentially, the S-pol antenna will be rotated at 6 to 7 deg/sec in order to keep the standard error of the estimates of the polarimetric observables (e.g., Zdr, rhoHV, phidp, LDR) at acceptable levels. On the other hand, the TOGA antenna will be rotated at 18-21 deg/sec. This will be true for all scanning scenarios. For more details on the characteristics of the two radars, see the Appendix below.
After the rain-surveillance scans for TOGA and S-pol are completed, there are 11.5 minutes left in each 15 minute block to accomplish the remaining scientific goals of TRMM-LBA Brazil. Each of the three scanning modes identified above will utilize this remaining time differently.
The "case study" scanning mode will be used most frequently during the field campaign because it will provide the necessary temporal and spatial resolution to accomplish the primary scientific goals listed above (particularly a-c). In this mode, dual-polarimetric, dual-Doppler, aircraft, lifecycle, and/or vertical structure studies will be accomplished of an individual convective system or cell as appropriate. For these type of studies, it is imperative to maximize the spatial and temporal resolution allowable by the radars.
Both the TOGA and S-pol radars will scan in sectors (typically 90 to 180 degrees wide). This will allow scan times to be from three to seven minutes long depending on the horizontal and vertical extent of the convective system and the range to the target. Horizontal resolution is better than 1 km at ranges less than 60 km from the radar (as defined by the 3dB beamwidth and azimuthal spacing which are both 0.93 degrees). The vertical resolution is specified as better than 1km for ranges less than 60 km and temperatures greater than -20 degrees Celsius (or heights below about 8 km AGL). Above 8 km AGL, the vertical resolution is set at better than 2km within 60 km range. The case study scans should be configured to top all convection of interest and obtain full horizontal coverage. However, it is recognized that storms at close ranges (< 25 km) can pose serious problems for these goals. Our philosophy in this situation is to attempt to scan all "significant echo" as defined by the application. For example, minimum requirements for dual-polarimetric studies would include horizontal coverage of "significant echo" (e.g., Zh>10-30 dBZ) up to the top of the mixed phase zone within 5-7 minutes or less. Dual-Doppler studies are more restrictive in the vertical, requiring full echo coverage to obtain a profile of divergence. Aircraft studies will necessitate the focused scanning of cells being sampled at high temporal resolution.
Using these specifications and a "back-wall" scan optimizer, a "case study" scan template was created (Fig. 6). Note that we plan for a minimum cone of silence at a range of 15 km of 8 km in height (T=-20C). Operations within 15 km in range will be severely limited in echo coverage and/or temporal sampling. This scan will be programmed into the S-pol and TOGA radars for use in the case study and mesoscale study modes. The user will merely choose the azimuthal limits (i.e., sector size and location) and maximum elevation angle (or maximum scan time as an alternative).
For example, Fig. 7 (Fig. 8) depicts the scan time vs. maximum elevation angle for the SPOLSec (TOGASec) scan type as a function of sector size. In case study mode, S-POL (TOGA) will be rotated at 7 deg/sec (18 deg/sec). For more details on the scan types or on the radars themseleves, see the Appendix below. Note that sector times, the maximum elevation angle (or maximum height as can be determined from Fig. 6), and sector size are all competitive characteristics. The radar scientist(s) on duty will have to determine the priority for each of these constraints depending on the application.
A typical series of scans for both radars during the case study mode is shown in Fig. 9. Of course, both radars begin by executing a 3.5 minute long rain-surveillance scan (see above). Each radar then performs a series of sector and RHI (Range Height Indicator) scans to fill the remaining 11.5 minutes. The goal in this example is to keep the slower rotating SPOL scans nearly in sync with the TOGASec scans (typically for the purpose of dual-Doppler analyses). To accomplish this task, the S-POL (TOGA) radar scans a 120 (180) degree sector with a maximum elevation angle of 24 degrees (37 degrees). High vertical resolution RHI scans are utilized to better characterize the vertical structure of the convection. The RHI's are strategically placed to keep the radars nearly synchronized. Of course, the actual series of scans, sector sizes, scan times, and maximum elevation angles will depend on the application. This example can serve as a template for combined dual-Doppler/dual-polarimetric studies of deep convection within the dual-Doppler lobes. In this area, the radars could scan convective systems with tops of 12 - 20 km and horizontal extent of 60 - 200 km in 4 - 5 minutes. Using the scan templates defined by Fig. 6 - Fig. 8 (and Table 2), the radar operator can easily customize this "case study" series of scans.
The mesoscale scanning mode will be used for widespread convective echoes at close range, mesoscale statistical studies, TRMM satellite/ER-2 overpass flights, and general surveillance. When there are not specific cells targeted for "case study" type research, both radars will execute full volume (360 degree) sector scans. Scans for both the TOGA and S-POL radars will follow the template shown in Fig. 6. The S-POL radar will be rotated at 7 deg/sec for the full 360 degrees (see Fig. 7 for the time vs. elevation angle plot). The TOGA radar will be spun slightly faster in 360 degree mode (21 deg/sec) as compared to sector mode (18 deg/sec). This will allow the TOGA radar to sample its surroundings in shorter time and still obtain good estimates of reflectivity and radial velocity. See Fig. 10 for a plot of time vs. maximum elevation angle for the "TOGA360" scan.
An example series of scans for the mesoscale study scanning mode is presented in Fig. 11. As before, both radars execute a low-level 360-degree volume scan for mapping rainfall. In the remaining 11.5 minutes, the S-POL radar will measure the polarimetric variables up to a maximum elevation angle of 16 degrees for the full 360 degree volume. The faster rotating TOGA radar will accomplish two full volume scans up to 24 degrees in elevation and have 1 minute left over for RHI studies of vertical structure in the same time frame. This scan mode will be ideal for gathering mesoscale polarimetric radar statisitics over Rondonia. In addition, it will be used for general surveillance and situations with widespread convective echo (assuming no single system is of interest to other operations). Finally, this scan mode will be used with coordinated ER-2 flights during TRMM satellite overpasses in order to provide ground validation data for initial satellite algorithms. Since the ER-2 overflights will cover a significant horizontal distance while sampling the mesoscale convection below with its remote sensing instruments, a full 360 degree scan strategy is the best way to provide complete ground validation data.
Although a low-level, full volume rainfall scan will be executed every 15 minutes during the field campaign, there will be instances when focused rainfall studies will be accomplished. For example, when there is significant rainfall over one of the rain gauge networks ("close" network with disdrometer and profiler or "far" network; see Fig. 2), we plan to obtain high temporal and spatial resolution data over the area using tight sectors. In order to investigate the physics and assumptions in the GV efforts, the S-POL radar will obtain one minute updates over the gauge networks with 250 to 500 m vertical resolution up to 4 km AGL. To accomplish this task, S-POL will rotate at 6 deg/sec over a 25 degree sector centered on the gauge networks.
An example of the height vs. range position of the sweeps is given in Fig. 12 for the near network (R=35 km to the northwest of S-POL) "GaugeSec" scan. Fig. 13 presents the same information for the far network (R= 90 km to the north of S-POL). The one minute temporal resolution of the S-POL data will match that of the tipping bucket gauges. Also note that the first sweep (first two sweeps) for the near (far) GaugeSec scan is partially blocked on purpose. We hope to take advantage of the fact that Kdp (the specific differential phase) is nearly immune to the effect of partial beam blockage. This will allow us to obtain rainfall measurements closer to the ground (particularly for the "far" network), thus minimizing the typical problems of radar rainfall measurements (e.g., horizontal advection, evaporation, and drop size evolution).
An example of the series of scans for both the TOGA and S-POL rain gauges is given in Fig. 14. The S-POL follows the low-level, 360 degree rain scan with a series of ten (10) rapid sector scans over the appropriate gauge network. The S-pol series is completed with several RHI scans to document the overall vertical structure over the gauges. Meanwhile, the TOGA radar will repeat the RainSurT scan (low-level, 360 degree rain scan) four times and end with a few RHI's. In effect, The TOGA radar will measure the mesoscale rainfall pattern while the S-POL radar is focused exclusively over the rain gauge network.
Please direct any questions, comments, or suggestions to Larry Carey at