High-Resolution RUC CAPE Values and Their Relationship to Right Turning Supercells

Transcription

1 High-Resolution RUC CAPE Values and Their Relationship to Right Turning Supercells ANDREW H. MAIR Meteorology Program, Iowa State University, Ames, IA Mentor: Dr. William A. Gallus Jr. Department of Geological and Atmospheric Sciences Iowa State University, Ames, IA ABSTRACT Supercell movement is very important to severe weather prediction. Without knowing the direction a supercell will turn forecasting where one will go and how to issue warnings becomes much more difficult. In this study, data were collected from supercells during June 2008 across the United States. The 13 km high resolution RUC model was used to find a relationship between turning supercells and CAPE. The results show that CAPE values decrease 1-2 hours before a supercell right turns. The results also show that supercells that turn have a high CAPE value before compared to supercells that maintained a straight path. 1. Introduction Supercell movement is very important to severe weather prediction. Surprisingly very little research has been conducted on right turning supercells. Most of the research has focused on the new direction a supercell would travel rather than when the cell will turn. The focus of this research is to try to find a link between when a supercell right turns with CAPE values. CAPE stands for convective available potential energy, which means how much energy is available for convection when a parcel is lifted through the atmosphere. My original hypothesis formed was a supercell will become a right turning supercell due to mesocyclone strengthening. Therefore, CAPE should increase when a supercell right turns because CAPE has been show to strengthen the mesocyclone. Most supercell research seems to be focused on the degree to which a supercell will right turn rather than when it will turn. The most current method of determining right turning supercell movement is Bunkers et al. (1999) hodograph technique. In this method, the supercell movement is determined by looking at the hodograph and calculating the motion the storm will take if it right turns (Bunkers et al. 1999). There are a few older methods for predicting supercell motion, but Bunkers et al. (1999) research is the most up-to-date and statistically the most accurate theory for supercell motion. his newest theory is more accurate. The previous method focused on calculating the mean wind, followed by decreasing the speed by 25% and the motion changing 30 degrees (Maddox, 1975). Another older method involves calculating the mass flux between the updraft and downdraft of the storm (Colquhoun, 1980). What makes supercell motions very hard to predict is that right turning

2 supercells tend to go against the mean flow of the winds (Keer and Darkow, 1995). It has been observed in the meteorology community that nonsupercell storms move with the mean flow of the storm. However, storms that are stronger and live longer tend to turn to the right of the mean wind and slow down, this is where the term right turner comes from. Very little research has actually been covered on the timing of supercell turning. No current method exists for determining the timing of a storm turning. Most, if not all, of the techniques for supercell motion discussed tend to focus on the radiosonde and hodograph. This is not useful when trying to determine when a supercell will right turn due to the limited areas of where radiosondes are released. A factor that may affect a supercell right turning is the cold pool (Bunkers et al. 1999). There are also outside influences on a supercell and right turning, such as orography (Bunkers et al. 1999). Bunker s paper states that this may have a large influence on supercell movement because of the increased or decreased convergence caused on the leeward and windward side of mountains. Storms have been witnessed to stall or redevelop causing difficulty on supercell motion (Bunkers et al. 1999). 2. Data and methods Data was collected from June June was chosen because it was the most active month during the 2008 season. Data from only one year was looked at to ensure no change in resolution of the RUC analysis occurred. In order to find cases that fit the study, storm reports and reflectivity data were observed from the UCAR image archive. Data were selected when storm reports showed a tornado report and showed supercell characteristics on the radar image archive. Gibson Ridge Level 2 Analyst Edition radar files were then downloaded from NCDC. Using Gibson Ridge, latitude and longitude were recorded every half hour. Data was collected from anywhere in the United States, however, most of the data came from the central plains, where a total of 82 cases were collected with 54 supercells that followed a straight path and 28 cases that right turned. a. Storm initiation and dissipation In order for the storm to be initialized, radar reflectivity must show a return of at least 40 dbz. The storm was then followed until dissipation, which was when reflectivity returned below 40 dbz, or when the supercell becomes merged with another storm. Data was recorded for both storms that turn and storms that follow a straight vector. b. Track and turning In order to determine if a storm turns or if it is considered to be moving straight, a 20 turn from its current vector must be observed in the storm. Normal convention for supercells turning is a 30 turn, but 20 was used to account for errors in storm location. Also, errors in latitude and longitude caused by being distant to the radar can have an effect on correct supercell locations. Storm motions were calculated by using the GR2 Analyst storm motion tool. The storm was observed and where it appears to make a turn storm motion was calculated before and after and then compared. If a 20 turn in direction was observed then the storm is considered right turning. c. Collecting CAPE values In order to collect CAPE values, files were downloaded from the NCDC NOMADS site. The files were then converted and analyzed in Grads. Using 13 km high resolution RUC output CAPE values were recorded for the supercell at its current latitude and longitude on the hour. When collecting a CAPE value on the half hour, CAPE values were recorded for the hour before and the hour after and then averaged to obtain an interpolated CAPE value for that location. Surface based CAPE was used due to supercells being relatively low in the atmosphere and surface based CAPE would have the largest impact in a developed supercell. d. Statistical Analysis Storms were separated by those that turned and those that stayed on a straight path. Averages were then calculated and then

3 compared. A comparison was made between CAPE in the first hour of a storm s life and the last hour for both right turners and straight storms. June 7 th and 18 th were looked at for individual case patterns for individual storm data. June 7 th and 18 th were selected due to the largest amount of individual storms during June a supercell s life. However, when looking at individual storms, CAPE values fluctuate widely and change differently. 3. Observational Trends For this study, CAPE was observed for all supercells that either maintained a straight path or storms that made a right turn. A right turn is defined for this study as a turn to the right of the storm vector of at least 20. a. Non-Turning Supercell Trends For a supercell to be considered a nonturner, it has to have a storm vector that did not deviate to the right more than 20. Left turning storms were ignored for this study, so all non turning supercells maintained a straight path. Overall, data from non-turning supercells was very diverse. Some supercells had very low CAPE values while others had over 4000 J/kg of CAPE. As shown in Figures 1 and 2, even when only looking at data from one day, values are very diverse. Figure 2. CAPE values of non turning supercells from June 18 th 2008, throughout the day. b. Right Turning Supercell Trends Trends for right turning supercells are much more apparent than non-turners. As seen in Figure 3, CAPE values are much larger for the first 2 hours of a storm s life and then begin to fall approximately 2 hours before the storm right turns. After right turning, the supercells CAPE remains almost constant. Figure 1. CAPE values of non turning supercells from June 7 th 2008, throughout the day. On average, when a non-turner initializes, CAPE values are around 1351 J/kg, and when they dissipate they have a cape around 1215 J/kg, resulting in a 136 J/kg CAPE drop through Figure 3. CAPE values of turning supercells in June Storms were lined up so that all right turning occurred at the same time on the plots. Values for right turners are still very diverse as to be expected, but trends definitely show the values decreasing as the storm gets close to making its turn. Figure 4 shows box plots of the

4 values for every right turning supercell observed. As can be seen in figure 3, the straight line is the average CAPE throughout every recorded value. The varying blue line shows the average CAPE for that time period. As the storm develops, the values stay above the average, but as the storm gets closer to right turning, the values drop below the average. The values never return until the very end of storms lifetime. The last value can be seen as negligible due to very sparse data being recorded 4 hours after a storm right turned. The average CAPE value for initiation of right turners is 1953 J/kg, while at dissipation the average was CAPE was 1220 J/kg. CAPE at the time of turning was approximately 1525 J/kg. Thus, right turning supercells experienced an average CAPE drop of 733 J/kg, and on average 428 J/kg occurred before making a right turn. By looking at Figures 3 and 4, most of the CAPE dropped approximately 2 hours before turning. Figure 5. Shows the CAPE values of right turning supercells from June 7 th All storms are shown so that right turning takes place at the same time on each storm. In Figure 6, a similar trend appears. On 2 of the 3 storms, the CAPE decreased before the storm right turned. The supercell that did not have a CAPE drop before turning does, however, show a CAPE decrease after. Figure 4. CAPE values of all turning supercells in June Also, displays box plots to show variability in data. The linear light blue line is the average for all data collected, and the black line is the mean for that time period collected. Individual plots were also looked at for individual storm trends. Figure 5 shows all right turning supercells for June 7 th. Of the 5 right turning storms observed that day, 4 of the 5 decreased in CAPE approximately 1 to 2 hours before turning, showing consistency with the averaged values obtained. Figure 6. CAPE values of right turning supercells from June 18 th, All storms are shown so that right turning takes place at the same time on each storm. One possible reason CAPE decreases during right turning supercells is the diurnal cycle of CAPE each day. Figure 7 shows this cycle from a few cities during clear sky events in June. From Figure 7, it can be seen that CAPE values don t begin to fall until 0z. CAPE also tends to max out around 19z. Of the 28 right turning supercells observed 22 of them turned before 0z.

5 Figure 7. Diurnal cycle of CAPE change by hour. Data collected was from clear skies in 4 cities near most supercells. Data was also collected from June 7 th, in an area that experienced clear skies that day. Statistical analysis was performed in Figure 8 showing a projected change in CAPE 1 hour before turning until turn time. Confidence intervals were used and show the statistical odds of a right turning supercell that is identical to the ones collected having a change of CAPE in the range given. Figure 8. Confidence test of projected CAPE change from 1 hour before right turning until time of turn. A T-test was also performed on the orgininal hypothesis and the null hypothesis was unable to be rejected. Therefore, this does not support the original hyothesis of supercells gaining CAPE when right turning. 4. Comparisons of Turning and non Turning Supercells Right turning supercells show a much more organized pattern of CAPE tendencies than nonturning supercells (see Figures 1 and 2). Nonturning supercells seem to fluctuate largely compared to turning supercells. Figure 9. Average CAPE values for both turning and non turning supercells for the first 1.5 hours past initialization and the last 1.5 hours until dissipation. In Figure 9, you can see that right turning supercells on average had much higher CAPE values to begin with and tend to fall throughout the storms life. When looking at non-turning supercells, the CAPE values seem to consistently fluctuate throughout the storms life. Therefore, right turning supercells tend to have a much larger drop in CAPE compared to nonturning supercells. Although when looking at individual storms, many non right turning supercells showed a CAPE drop during the storms life, shown in Figures 1 and 2. Thus, concluding that a CAPE drop only occurs in right turning supercells cannot be made. However, 19 of 27 right turning supercells did show a CAPE drop before turning. One common trend that can be observed when comparing Figures 1 and 2 to 5 and 6, is how much more consistent right turning supercells are compared to non-turners. Most values on each cell do not change more than 1000 J/kg of CAPE. However, when looking at Figures 1 and 2, the CAPE values changes drastically over a short time period and have less stable CAPE values.

6 Figure 10. Averaged data from June 7 th, Data was averaged for turning supercells before turning, turning supercells after turning, and non turning supercells. Figure 10 shows a trend that can be observed. CAPE values before the supercell right turns are higher than non right turning supercells, but when looking at post-storm turning CAPE, CAPE values seems to become much smaller. At the end of the time period the values seem to begin to move in unison and decrease at the same rate. This would indicate that the whole atmosphere is beginning to become more stable and that CAPE values are falling consistently. It can be concluded that supercells that turn tend to have higher CAPE values before turning. This trend can also be observed in Figures 9 and 10 where CAPE values are much higher in turning supercells compared to non-turning supercells. The average CAPE value of the first hour of right turning supercells is 1927 J/kg, while nonturning supercells only averaged 1315 J/kg. So as a result right turning supercells averaged 612 J/kg of CAPE more than non turning supercells. 5. Conclusion Based on the dates shown above, it can be seen that right turning supercells tend to initialize with higher CAPE values and approximately 1 to 2 hours before turning, tend to show falling CAPE values. The CAPE values after the storm turns tend to be even lower than the CAPE of non-turning supercells. Unfortunately, many individual cases of non turning supercells show CAPE values in the same range as turning supercells. As a result, it cannot be concluded that all turning supercells show higher CAPE values all the time, but rather frequently have larger values. Non turning supercells tend to not show any consistent pattern with CAPE values. They tend to fluctuate much more than that of right turning supercells which indicates that other parameters need to be considered. My original hypothesis seems to be the opposite of what was observed. My original hypothesis was that CAPE values would rise when a supercell right turns. The direct opposite seems to have occurred. Turning supercells seem to lose CAPE approximately an hour to 2 hours before turning. Further research needs to be done on other parameters that strengthen a supercell s mesocyclone. In particular, helicity or other wind fields that assist in strengthening a mesocyclone. 6. Acknowledgements I would like to thank Dr.Gallus for his assistance on this project. I would also like to thank Jeff Duda for his help on running and using Grads, as well as Dave Flory for some procedure assistance, Justin Schultz for his guidance, and Ryan Alliss for statistical analysis. REFERENCES Brandes, 1977: Mesocyclone Evolution and Tornadogenesis: Some Observations. Monthly Weather Review 106, Bunkers, Klimowski, Zeitler, Thompson, and Weisman: Predicting Supercell Motion Using a New Hodograph Technique. Weather and forecasting 15, Colquhoun, J. R., 1980: A method of estimating the velocity of a severe thunderstorm using the vertical wind profile in the storm s environment. Eighth Conf. on Weather Forecasting and Analysis,

7 Denver, CO, Amer. Meteor. Soc., Donaldson, and Desrochers 1989: Improvement of Tornado Warnings by Doppler Radar Measurement of Mesocyclone Rotational Kinetic Energy. Weather and forecasting 5, Kerr, and Darkow, 1996: Storm-Relative Winds and Helicity in the Tornadic Thunderstorm Environment. Weather and forecasting 11, Maddox, 1976: An Evaluation of Tornado Proximity Wind and Stability Data. Monthly Weather Review 104, Richard, and Edwards, 2000: An Overview of Environmental Conditions and Forecast Implications of the 3 May 1999 Tornado Outbreak. Weather and forecasting 15,