Tropical Storm Hermine: Heavy rainfall in western Gulf By Richard H. Grumm National Weather Service Office State College, PA 16803 1. INTRODUCTION Tropical storm Hermine, the eighth named tropical system of the 2010 hurricane season, formed in the Bay of Campeche on 6 September 2010. The storm produced heavy rainfall (Fig. 1) as it ambled northward into Texas (Fig. 2). The satellite rainfall data show rain fell on 3-4 September (Fig. 1b) with the incipient circulation and then the rainfall area moved northward as the surface circulation moved northward (Figs. 1a-d). The surface low moved over Texas on the 6 September and brought heavy rainfall (Fig. 2) and severe weather, mainly in the form of tornadoes, to both Texas and Oklahoma (Fig. 3). As late as 4 September, the potential tropical development in the Gulf was considered a low probability event. The NCEP models and SREF showed the potential for heavy rainfall from forecasts issued on 4 September. The rainfall was associated with a surge of high precipitable water air moving onshore with a weak low. This paper will summarize the pattern and the heavy rainfall associated with tropical storm Hermine. The focus is on the patterns and anomalies that produced the heavy rainfall. Forecasts from the NCEP SREF are the focus of the rainfall forecasts. 2. METHODS The pattern was reconstructed used the NCEP GFS and NAM. All data were plotted in GrADS (Doty and Kinter 1995). The higher resolution NCEP NAM is used to show the conditions during the event. The anomalies were computed from the NCEP/NCAR re-analysis data (Kalnay et al 1996) as describe by Hart and Grumm 2001 and Grumm and Hart 2001. Unless otherwise stated, the base data was the GFS or NAM and the means and standard deviations were computed by comparing the NAM to the NCEP/NCAR 30-year climatological values. For brevity times are referred to in the format of 08/1800 for 08 September 2010 at 1800 UTC. 3. METHODS i. Large scale pattern The large scale pattern over the United States from 3-6 September is shown in Figure 4. The key features include the retrograding ridge over Florida and southwestern Atlantic and the circulation with Hurricane Earl off the East Coast of the United States (Fig. 4a-c). Hermine would form in the weakness or tropical upper tropospheric trough (TUTT) over the western Gulf (Figs. 4a-f). The surge of high PW air about the subtropical ridge, associated with what became Hermine is evident in Figure 5. The high PW air, with 3-5 SD PW anomalies moved from the Bay of Campeche into eastern Mexico then into Texas (Fig. 5a-i). This surge of high PW air was closely aligned with the heavy rainfall.
iii. Rainfall analysis ii. Regional pattern The GFS analysis showed a circulation over the western Gulf at late on 5 September (not shown). By 06/1200 UTC (Fig. 6a) the GFS picked up a weak circulation south of Texas. The negative pressure anomalies highlight this feature well. The NAM analysis was strikingly similar and is not shown. The NAM PW and anomalies is shown in Figure 7. The NAM PW values were slightly higher at times than the GFS (not shown). Though both systems showed high PW values (60-65 mm) and large PW anomalies. The NAM PW anomalies peaked in the +4 to +5SD range over the Gulf (Figs. 7a-c) and over Texas (Fig. 7d-e). The NAM closed of a 75mm PW contour at 07/1200 UTC (Fig. 7e). The GFS never showed a contour (in 5mm increments) greater than 65 mm. In addition to the anomalously high PW values, the circulation with Hermine had strong low-level winds. The 850 hpa winds showed strong winds and 2-4SD wind anomalies come onshore around 06/1800 UTC (Fig. 8b) ahead of the main circulation center. The strong winds with the actually circulation center cam into Texas around 07/0600 UTC. The 850 hpa total wind anomalies peaked near +6SD above normal in southern Texas at 07/1200 UTC (Fig. 8e). The heavy rainfall axis lined up well with the stronger southerly winds and the implied convergence between those winds and the strong southeasterly winds. The 850 hpa moisture flux (MFLUX) in Figure 9 highlights this effect. MFLUX anomalies peaked at over 6SD above normal in the regions where the heaviest rains were observed (Fig. 9f-i) and Figure 2. The satellite rainfall data over the Gulf with the initial development and surge with Hermine was shown in Figure 1. The rainfall over Texas for the event was shown in Figure 2. These data showed heavy rains near the coast with 96 to 192 mm of in that region. Inland a large area of 128 mm or more rainfall was present along a southwest to northeast axis. Maximum rainfall in the coarse UPD data was 256 mm, about 10 inches of rainfall. There was also a large area within 192 mm contour. This storm produced some hefty rainfall amounts. Figure 10 shows the rainfall in 4 discrete periods ending at 1200 UTC 6-9 September. The rainfall along the coast fell mainly on the 6 th and the inland rainfall fell mainly on the 7 th and 8 th (Figs. 10c-d). Clearly the period of 07/1200 through 08/1200 UTC was the biggest rainfall period over the largest area of Texas. This matched well with the surge of high PW air and high MFLUX values (Figs. 7 & 9). iv. Forecasts Forecast of QPF from the NAM, GFS and SREF are presented in Figures 11-13. The patterns were relatively well predicted and thus the precipitation fields were also relatively well predicted. The NAM forecasts from as early as 04/1200 UTC showed an axis of heavy rainfall in Texas (Fig. 11a). Clearly, the NAM was too slow and as the forecast length decreased (Figs. 11a-i) the NAM moved the axis of heavy rainfall inland. The long range 04/1200 UTC forecast only has data through 08/0000 UTC as the NAM is only run out to 84 hours. Thus, the NAM from this time has no data after 08/0000 UTC.
Interestingly, some of the shorter range forecasts may have moved the rainfall too far west. The lower amounts in the 07/0600 and 07/1200 UTC data (Figs. 11h-i) will not show heavy rainfall that already fell. Thus, comparing these data to Figures 2 & 10 should be done with caution. The comparable GFS forecasts are shown in Figure 12. The overall pattern is quite similar to the pattern shown by the NAM. The details clearly differ but the pattern and concept of heavy rainfall near the coast than an axis of near 4 inches of rainfall (96 mm) with orientation similar to what was observed is quite evident. Figure 13 shows SREF QPF s from 9 runs. These data show the ensemble mean QPF which will wash out maximum values and not show the range or uncertainty. Forecasts shown end at 08/0000 UTC the range of the shortest forecast. Figures 14 & 15 address these issues. But these data show that forecasts from as early as 04/0900 UTC had heavy rainfall moving into southern Texas (Fig. 13a). These data do not cover the entire time shown in the GFS and NAM images. Both forecasts did show heavy rainfall moving into coastal regions. Figures 14 & 15 show the SREF probabilities for discrete times and thresholds. The 04/0900 UTC forecasts (Fig. 14) showed that heavy rain was most probable on the 7 th. The near 100% chance of 35 mm or more QPF in the 30 hour period ending at 07/2100 UTC was a good forecast, as was the closed 64 mm contour indicating locally heavy rainfall. Forecasts initialized 2 days later (Fig. 15) showed the rainfall farther inland on the 8 th. The probabilities are a bit lower but both the pattern of the mean rainfall and probabilities got the general area correct. 4. CONCLUSIONS Tropical storm Hermine developed in the Bay of Campeche on 6 September and moved northward. The storm produced rain over the western Gulf (Fig. 1) and portions of Mexico before providing heavy rainfall and tornadoes to Texas (Figs. 2 & 3). The circulation with Hermine clearly brought strong winds, high PW values with high PW anomalies into Texas. The heavy rainfall areas were relatively well defined by the area of above normal PW air and the strong lowlevel 850 hpa winds (Figs. 7 & 8). The 850 hpa winds were near 6SDs above normal at times in the region of the heaviest rainfall. The heaviest rainfall over Texas was well aligned with the areas of high MFLUX (Fig. 9). MFLUX anomalies were slightly higher than 6SDs above normal at times near the areas where the heaviest rain was observed. The crude UPD data suggested that a large area of east-central Texas received over 5 inches of rainfall and up to 10 inches of rain was observed locally. Forecasts suggest that the surge of high PW air and thus the potential for heavy rainfall was captured by GFS, NAM, and SREF forecasts initialized at 4 September 2010. Early forecasts were not shown here. But the signal and high QPF potential were well predicted in the NCEP guidance. 5. Acknowledgements 6. References Doty, B. E., and J. L. Kinter III, 1995: Geophysical data and visualization using GrADS. Visualization Techniques Space and Atmospheric
Sciences, E. P. Szuszczewicz and Bredekamp, Eds., NASA, 209 219. Doty, B. E., and J. L. Kinter III, 1995: Geophysical data and visualization using GrADS. Visualization Techniques Space and Atmospheric Sciences, E. P. Szuszczewicz and Bredekamp, Eds., NASA, 209 219. Grumm, R.H. and R. Hart. 2001: Standardized Anomalies Applied to Significant Cold Season Weather Events: Preliminary Findings. Wea. and Fore., 16,736 754. Hart, R. E., and R. H. Grumm, 2001: Using normalized climatological anomalies to rank synoptic scale events objectively. Mon. Wea. Rev., 129, 2426 2442. Junker, N. W., R. H. Grumm, R. Hart, L. F. Bosart, K. M. Bell, and F. J. Pereira, 2008: Use of standardized anomaly fields to anticipate extreme rainfall in the mountains of northern California. Wea. Forecasting,23, 336 356. Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40- Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77,437 471. Onogi, K., J. Tsutsui, H. Koide, M. Sakamoto, S. Kobayashi, H. Hatsushika, T. Matsumoto, N. Yamazaki, H. Kamahori, K. Takahashi, S. Kadokura, K. Wada, K. Kato, R. Oyama, T. Ose, N. Mannoji and R. Taira (2007) : The JRA-25 Reanalysis. J. Meteor. Soc. Japan,85,369-432. Lin, Y. and K. E. Mitchell, 2005: The NCEP Stage II/IV hourly precipitation analyses: development and applications. Preprints, 19th Conf. on Hydrology, American Meteorological Society, San Diego, CA, 9-13 January 2005, Paper 1.2.
Figure 1. Climate Prediction Center Morphed (CMORPH) rainfall estimates (mm) for the 24-hour periods at ending at 0000 UTC a) 4 Septebmer, b) 5 September, c) 6 September and d) 7 September 2010.
Figure 2. Total rainfall from the unified precipitation data set (mm) focused over the southern United States from 1200 UTC 5-8 September 2010.
Figure 3. Storm reports by event type showing severe weather over the United States on 7 & 8 September 2010.
.. Figure 4. GFS 00-hour forecasts showing 500 hpa heights (m) and 500 hpa height standardized anomalies (sigma) valid at a) 1200 UTC 03 September, b) 0000 UTC 04 September, c) 1200 UTC 04 September, d) 0000 UTC 05 September, e) 1200 UTC 05 September, f) 0000 UTC 06 September, g) 1200 UTC 06 September, h) 0000 UTC 07 September, and i) 1200 UTC 7 September 2010. Return to text.
Figure 5. As in Figure 4 except showing precipitable water (mm) and precipitable water anomalies. Return to text.
Figure 6. As in Figure 4 except GFS 00-hour forecasts of mean sea level pressure (hpa) and pressure anomalies in 6-hour increments from a-i) 6-8 September 2010.
Figure 7. As in Figure 6 except for NAM 00-hour forecasts showing precipitable water and precipitable water anomalies. Return to text.
Figure 8. As in Figure 7 except for NAM 850 hpa winds (kts) and wind anomalies. Return to text.
Figure 9. As in Figure 8 except for NAM 850 hpa moisture flux and moisture flux anomalies. Return to text.
Figure 10. As in Figure 2 except for rainfall in 24 hour increments valid for the 24 hour periods ending at 1200 UTC a-c) 6-9 September 2010. Return to text.
Figure 11. As in Figure 9 except for NAM forecast of total QPF valid for the time period ending at 1200 UTC 09 September 2010. Forecasts were initialized at a) 1200 UTC 4 September, b) 1800 UTC 5 September, c) 0000 UTC 6 September, d) 0600 UTC 6 September, e) 1200 UTC 6 September, f) 1800 UTC 6 September, g) 0000 UTC 7 September, h) 0600 UTC 7 September, i) 1200 UTC 12 September 2010. The 4 September run has no data forecast after 0000 UTC 8 September 2010. SepReturn to text.
Figure 12. As in Figure 11 except for GFS forecasts. Return to text.
Figure 13. As in Figure 12 except showing 9 SREF members mean QPF (mm) valid at 00000 UTC 08 September 2010. Data are from SREF forecasts initialized at ) 0900 UTC 4 September, b) 1500 UTC 4 September, c) 2100 UTC 4 September 2010, d) 0300 UTC 5 September, e) 0900 UTC 5 September 2010, f) 1500 UTC 5 September 2010, g) 2100 UTC 6 September 2010, h) 0300 6 September and i) 0900 UTC 6 September 2010. Back to text.
Figure 14. NCEP SREF forecasts initialized at 0900 UTC 4 September valid at 2100 UTC 7 September 2010 showing a) probability of 12mm in the past 6 hours, b) probability of 25 mm in the past 24 hours, c) the probability of 35 mm I the past 30 hours and d) the probability of 50 mm in the past 36 hours. Contours show ensemble mean QPF in mm. Return to text.
Figure 15. As in Figure 14 except for showing forecasts initialized at 0900 UTC 6 September. Return to text.