1. Polarimetric radar analysis of convection in the complex topography
of northwestern Mexico during NAME 2004
Timothy J. Lang, Steven A. Rutledge, Robert Cifelli, Lee Nelson, and Stephen W. Nesbitt
1. Overview
3. Statistical Analyses
4. Case Study Analyses
In this poster we report on preliminary results from an investigation into the characteristics of convection during NAME 2004, emphasizing rainfall and microphysical structure, as well as the effects of terrain.
While reflectivity statistics are similar among the different terrain bands, clear differences in D0 [D0 via D0 = 1.529*(ZDR)0.467] between land (coastal plain and SMO) and sea (Gulf of California) exist. Generally, larger D0s are more common over land, and for high reflectivities (i.e., convection) the sea D0 is typically smaller than the land D0.
A cell-tracking algorithm, based on TITAN and SCIT, has been applied to the ~35 case studies. Using this algorithm, the microphysical evolution of every significant cell can be examined in detail. Analyses can be broken down by terrain elevation band, etc.
Three radars in core Monsoon region (NAME Tier I; NW Mexico)
Network covered Gulf of California, Sierra Madre Occidental, Coastal Plain, Baja Peninsula, Pacific Ocean
S-Pol – S-Band, Polarimetric, Doppler
SMN – Cabo and Guasave radars
(C-Band, Doppler)
(only used S-Pol in this study)
Case study example
15 July 2004 (Case 05)
MCS over high terrain moved toward the coast. Analysis of two cells near their peak intensity reveals significantly different microphysical structures, with the coastal cell (Track 006) having substantial more graupel aloft than the higher-elevation cell (Track 008). Are these differences representative of more systematic variability with elevation?
NAME radar network (only Cabo, Guasave, and S-Pol data were available for analysis)
S-Pol 24-h Ops 7/8-8/21; Two Modes
Climatology
Most frequent; 200-km range
Full 360s, complete in 15 min
Rain angles (0.8,1.3,1.8°) & 0.0°
Storm Microphysics
~90 hours total over ~35 cases
Usually 150-km range
2-3 PPIs with 0-1 RHIs in 15 min
rain angles (0.8,1.3,1.8°)
On average, precipitating systems over the water contain lower precipitation-sized ice mass values (Cifelli et al. 2002) than those over land, while differences between the different land elevation bands are small .
Ice - solid
Liquid water - dash
2. Beam Blockage Correction
Blockage-correction and other and quality-control procedures followed Cifelli et al. (2002) and Lang et al. (2007).
0.8° (left) through 1.8° sweeps were blocked at S-Pol, often over 90%. There also sometimes were multiple blocks along a single ray.
Unblocked
Blocked
Temporal evolution of individual cells is also being examined. In this example (Track 022 from Case 05), precipitation ice mass (black contours; g m-3) aloft leads the fallout of liquid precipitation (blue contours) near the surface, and demonstrates the importance of ice-based precipitation processes in this cell.
For a given KDP range, ZH should also vary within a fixed range for rain (Scarchilli et al. 1996). Departures from this denote blocks. The magnitude of the mean departure is the ZH correction that must be applied.
Original KDP DEM
Land and ocean radar characteristics are nearly 12 h out of phase with one another. Convection peaks during the morning over water, during the afternoon over land. This is similar to the results of Lang et al. (2007) using the 2-D regional multi-radar dataset.
TRMM PR Intercomparisons
5. Preliminary Conclusions
KDP-based blockage correction demonstrates advantages relative to other methods, particularly for short-term deployments with limited raingage support (e.g., S-Pol during TIMREX).
The largest differences in precipitation and microphysical characteristics are between land and sea, not between different land elevation bands. Precipitation over the water is nearly 12 h out of phase with the land.
Precipitation over water appears to be the result of DSDs with smaller D0 values, consistent with reduced influence of ice-based precipitation processes.
Case studies are underway to study the detailed structure and evolution of NAME convection.
KDP and DEM-based corrections perform similarly well, with KDP in slightly better agreement with TRMM PR. Also, KDP is capable of correcting higher blockage than DEM (> 90%).
There are minor morning peaks in the land data. The afternoon peaks mostly coincide, with minor differences that do not follow a trend (sometimes low elevations peak first, sometimes high elevations do). This may be related to increased noise due to relatively fewer data points in any one land elevation band.
All TRMM overpasses mean offsets: KDP = dB; DEM = dB
Contact Info: Timothy Lang, CSU Atmospheric Science,
Ft Collins, CO 80523; (970) ,