A Forecast Procedure for Dry Potential

Nick Nauslar, University of Nevada, Reno

Timothy J. Brown, DRI

Jim Wallmann, National Service Reno WFO

Abstract

“Dry” (traditionally less than 2.5 mm or 0.1” of rainfall) forecasting has long been a forecast problem for the western United States. These dry are responsible for starting thousands of wildland fires every year. In the largest lightning outbreaks (or “busts” in the wildland fire community), hundreds of fires may be started in a 24 to 36 hour period. These extreme events put a huge strain on fire suppression efforts. Many of these fires may go unstaffed due to the lack of available fire personnel simply because of the large number of fire starts. Forecasting these events in advance, even just 24‐48 hours, could help fire agencies plan resources in preparation of a large outbreak. Fires are much more likely to be controlled during the early stages, and therefore cost much less to fight. Due to the seemingly innocuous conditions preceding dry thunderstorm development across the western United States (west of the Rocky Mountains), forecasting dry thunderstorm events can prove challenging and inconsistent. Jim Wallmann developed WA04, a conceptual model of dry thunderstorms that includes the pressure of the dynamic tropopause, jet streak dynamics, equivalent potential temperature, and upper level lapse rates in conjunction with the High Level Total Totals (Wallmann 2004). Isentropic analysis was added to WA04. This procedure was applied to two case studies from the of 2009 including the dry lightning event in northwest Nevada, northern California, and south central Oregon on August 1, 2009 and the east central Washington dry lightning event of August 20‐21, 2009. It proved useful in determining the potential for dry thunderstorm development in the preceding days and hours to the initiation of the event. This presentation summarizes these case studies, and describes the dry lightning forecast procedure.

1. Introduction 2008 northern California received over Dry lightning events or busts 5,000 lightning strikes that ignited create numerous fire starts in the western approximately 500 fires. This represented United States (west of the Rocky a top three lightning bust in that area Mountains), which stress local and throughout the history of the National regional initial attack resources. A Lightning Detection Network. Several of lightning bust refers to a system these busts occur annually mostly producing thousands of lightning strikes between June and September throughout over an area or region. During June 20‐21, the western United States. Many of these lightning busts occur with a trough the . Finally, the analysis of the approaching the western United States dynamic tropopause location and more that phases with monsoonal moisture importantly the transition or gradient of entrained from the south. Rorig and it in conjunction with the other Ferguson 1999 and Crimmins 2005 show parameters determines the potential for a short of synoptic patterns dry thunderstorm formation. for critical fire conditions in Pacific Other papers have aimed to Northwest and the Southwest United predict dry thunderstorms by States. However, it can be difficult to determining the importance of mid level diagnose the location of high‐based instability and low level dewpoint convection or to delineate dry and wet depressions (Rorig and Ferguson 2002 storm formations. These busts generate and Rorig et al. 2006). Rorig et al. 2006 problems not only for fire managers, but aimed to predict dry thunderstorms via to forecasters as well. Due to surface or an algorithm when convection was lower tropospheric thunderstorm indices expected by implementing thresholds (Totals Totals (TT), Convective Available based on moisture and instability Potential Energy (CAPE), K Index, Lifted components. Index (LI), etc.) not revealing the true The procedure detailed in this potential of most of these busts, many paper seeks to determine if dry events go under forecasted. To better thunderstorms will form and to some ascertain dry lightning potential, a extent, the intensity of the event. Two collaborative effort developed a cases chosen for this paper occurred 1‐2 procedure incorporating some non‐ August 2009 over northwestern Nevada, traditional thunderstorm parameters and northern California, and south central analyses. Oregon and 20‐21 August 2009 over The procedure includes isentropic central and eastern Washington. The first analysis, upper tropospheric level lapse case started in the late afternoon and rates (ULLR), high level total totals persisted through the evening hours and (HLTT), jet streak dynamics, and dynamic yielded over 6500 lightning strikes and tropopause analysis. Isentropic analysis approximately 200 fires throughout the entails the inspection of vertical cross‐ next week. The second case transpired sections with equivalent potential overnight with nearly 3000 lightning temperature and mixing ratio or potential strikes that included deployment of temperature and relative humidity incident management teams. These two plotted to determine instability and cases show the variability in the timing moisture plumes. Upper level lapse rate and the formation of these events, but values contoured illustrate the dry with the implementation of the procedure thunderstorm formation threshold of 7.5 both could be forecasted. The flexibility C/km. HLTT values of 25 C or higher and inclusiveness of the procedure for represent an environment conducive for different types of dry lightning events high‐based convection (Wallmann 2004. remains one of its strongest assets. These thresholds are based on empirical evidence and can be explained physically. 2. Procedure The influence and location of an upper level jet determine divergence aloft, Isentropic analysis has long been which aides the vertical development of used to ascertain east of the Rocky Mountains. southern Cascades case altered the upper Meteorologists also implemented it to a level wind speeds and direction enough to lesser extent for thunderstorm forecasts. induce divergence aloft and rising motion Isentropic analysis displays vertical below it. The jet does not have to motion in the atmosphere in the absence particularly strong either. On 20‐21 June of strong evidence at a constant pressure 2008 Northern California lightning bust, level. It also identifies moisture transport the polar and subtropical jets were not in a more complete manner. For these located over that region, but a smaller jet reasons, isentropic analysis provides a or jetlet existed between the two jets competent tool for distinguishing high‐ created in part by offshore convection. based thunderstorms, which occur This jetlet aided the upper level lift frequently within the intermountain west. throughout the event. Two sets of isentropic maps The dynamic tropopause is used to developed included contours of help assess the upper level forcing in two equivalent potential temperature (Θe) ways. First, weak short wave troughs can with mixing ratio and contours of be difficult to see at times on 500mb potential temperature with relative vorticity charts and therefore forecasters humidity (RH). By creating this type of may miss the weak forcing. Using the plot, one needs only to find the moisture dynamic tropopause on a potential plumes and instability (dΘe/dz). When vorticity surface, these weak short wave higher values of Θe are below lower troughs can be easier for a forecaster to values of Θe, this implies instability. If a locate as the trough will be represented plume or an increase of moisture by a tropopause undulation (Wallmann coincides with the instability, this 2004, and Hirschberg and Fritsch 1991a) becomes a focus for thunderstorm from which upper level vorticity development. To ascertain its dryness, advection and upward vertical motion can becomes more complicated. One can look be inferred. Second, the undulation is for lower RH or mixing ratio values below often associated with atmospheric warm this focus, and this can provide some air advection at that level which can support in deciding whether the storms induce near‐surface pressure falls with will be dry or wet. resulting low‐level convergence and The presence of an upper level jet upward vertical motion (Hirschberg and streak plays a vital role in the Fristch 1991a‐b). A large tropopause development of thunderstorms including undulation existed off the Oregon Coast dry thunderstorms. One main difference 20 June 2008 northern California case, from and dry but a weak fall in the tropopause moved thunderstorms is the jet divergent onto the California Coast at 18 UTC 21 region(s) do not have to be over the June. Significant lightning was associated development of dry thunderstorms. The with the trough as it moved onshore with influence of an approaching jet streak can numerous new fire starts. create enough mass momentum With bases often high during adjustments to allow for upper level dry thunderstorm outbreaks, sometimes divergence to couple with the other above 5 km above sea level, commonly procedural entities. An approaching jet in used instability parameters such as the 1 August northern Sierras and surface‐based lapse rates and the Total Totals index should be adjusted upward (Wallmann 2004 and Milne 2004). The important to note that while increasing high level total totals (HLTT – Milne values of HLTT and ULLR above the 2004) index and upper level lapse rates thresholds would suggest much better (ULR, 500‐300mb) are useful with values potential for thunderstorms the increase most conducive to extensive dry does not necessarily portend an increase thunderstorm development 30 C and 7.5 in the storms being dry. In fact, for HLTT C/km, respectively. For dry it is often the opposite as an increase in thunderstorms, HLTT can be as low as 20 the HLTT could be the result of higher C, but is most often found with values dewpoints at 700mb, which would result above 25 C (Wallmann 2004). It is in wetter storms.

Figure 1: Map of all lightning strikes throughout duration of 1‐2 August 2009 event

3. Case Study Analysis a) 1­2 August 2009

On 1 August 2009 a vigorous 200 fires from central Nevada to central trough approached the Pacific coast of Oregon (Figure 1). As with many of the California and Oregon. Ahead of this dry lightning busts, the event became trough, monsoonal moisture advected wetter with time. However, south central northwestward from the southwestern Oregon, northern California, and United States in the mid levels of the northwestern Nevada received several atmosphere. This set up resulted in over thousand lightning strikes with minimal 6500 lightning strikes and approximately precipitation.

Figures 2&3: 18 UTC 1 August 2009 and 00 UTC 2 August 2009 Θe (Kelvin, blue contours), mixing ratio (g/kg, colored filled with scale on left), and wind barbs (black) NARR cross‐ sections. Approximate location from southeast of Reno, NV to west central Oregon.

Examining the 18 UTC and 00 UTC increases again at 06 UTC (not shown). 2 August 2009 (Figures 2&3) North The moisture plumes indicate a American Regional Reanalysis (NARR) monsoonal push while the increase in cross sections that stretch from lower level moisture demonstrate the northwestern Nevada to western Oregon, effects of diurnal boundary layer a plume of higher mixing ratios and RH dynamics, and the storms’ precipitation exist in the 00 UTC cross section that did adding moisture below cloud base not exist in the 18 UTC cross section. securing the fact that this event became Additionally, the plumes of moisture wetter with time. collocate with potential/convective The moisture advected into this instability. The 336 K closed contour with region remains confined to below the the lower values above it indicates this midlevel of the atmosphere, and the instability. This occurs to a lesser extent plumes that aid in the convective further to the southwest where the 338 K development exist 700‐500mb region contour resides below lower values. (Figure 3). The atmosphere becomes very The lowest lower level mixing dry with height above these plumes. At ratio values reside below the greatest about 550mb the mixing ratio goes from 4 instability and moisture at mid levels g/kg to 1 g/kg at 425mb. This drier air showing a prototypical dry thunderstorm aloft can be cooled easier and thus event. This is in part due to afternoon helping to increase the ULLR. mixing and the lower level moisture

Figures 4&5: Contours of ULLR at 00 UTC and 06 UTC 2 August 2009. H’s and L’s represent the highs and lows of ULLR in the region.

At 00 UTC, the ULLR surpasses the remained southerly. It also increased the threshold of 7.5 C/km in along the path of wind the speed resulting in directional the cross section where most of the and speed divergence. While it seems lightning strikes occurred (Figure 4). At minimal on a synoptic scale, it is sufficient 06 UTC, the ULLR ranges from 7.8 C/km to provide enough divergence aloft to aid to 8 C/km in the region of interest (Figure lifting over this region. While the best 5). The collocation of the moisture, dynamics exist west of the Sierras, the isentropic instability, and sufficient ULLR best moisture resides east of them. reside along the path of highest density of Therefore, where the dynamics remain lightning strikes. The HLTT values near just sufficient enough and moisture exists, 30 C at Medford and Reno at 00 UTC the storms form and produced the supported the case for upward lightning bust during the evening of development. August 1st. The isentropic analysis The approaching polar jet from the distinguished greater atmospheric eastern Pacific aided the development of instability and moisture plumes that the lightning bust. At 21 UTC and 00 UTC otherwise might have looked more the jet streak within the larger 2‐cell jet is innocuous. The ULLR confirm that the oriented in a general southerly to upper levels support vertically developed northerly direction. While the area where thunderstorms, while the jet streak the storms formed were not in the analysis demonstrated mediocre but divergent region of this jet streak, and it sufficient divergence aloft. Finally, the dwelled on the fringe of the 2‐cell jet’s dynamic tropopause analysis indicated a divergent region. The approaching jet sharp gradient showing the transition provided a turning of the winds near the from a higher to lower pressure of the edge of the jet to southeasterly while the tropopause thus conveying cooler and winds in central Nevada and Oregon still drier air aloft approaching.

Figures 6&7: 00 UTC 2 August 2009 250mb Geopotential heights (black contours) and wind barbs (black with color gradient for speed). 00 UTC 2 August 2009 RUC analysis of 1.5 PV surface showing pressure of dynamic tropopause (green contours) and wind barbs (brown). b) 20­21 August 2009

On 20‐21 August 2009, a strong time with the strongest forcing east over trough and moved through Central Washington. By 09 UTC (Figure Oregon and Washington during the 9), the upper trough had progressed east evening hours. This trough triggered a into Central Washington with the nocturnal dry thunderstorm event over expected forcing over Eastern Central Washington that started several Washington and the Idaho Panhandle. new fires. The first thunderstorms The jet streaks associated with the formed around 0430 UTC 21 Aug and upper trough were also quite strong for a continued through 12 UTC before moving mid‐summer trough with peak winds in out of Eastern Washington (Figure 8). the core of the jets over 40 m/s. However, The initial thunderstorms were driest, the locations of the individual jet streaks before 08 UTC, and resulted in the in this case were not favorable for greatest number of problem fires with ageostrophic upward vertical motion. At two incident management teams 06 UTC (Figure 10), the leading anti‐ activated to coordinate the fight of the cyclonically curved jet streak was over largest fires. the Idaho Panhandle leaving Central For the event, the dynamic Washington in the left entrance region, tropopause was quite low west of the which would result in downward vertical Oregon and Washington Cascades at 06 motion. UTC, as low as 350mb, and sloped sharply Likewise, with the jet streak upward to the east. Strong upward moving onshore, the right exit region of forcing would be expected in the area to this jet was over Central Washington, also the east of the center of the trough at this supportive of downward vertical motion were near 7.5 C/km, which is at the due to the ageostrophic circulation. threshold of supporting dry convection while HLTT were above 30 C over Central Washington (not shown). Isentropic analysis along a cross‐section from Portland, Oregon to northern Idaho also shows the instability present east of the Washington Cascades. Over the east slopes of the Cascades, Θe values were maximized near 325 K above 700 mb with the isentropes decreasing slightly with height indicating and area of potential instability through at least 500 mb (Figure 12). In addition, the mixing ratio showed the largest amount of mid‐ level moisture in this region as the 4 g/kg Figure 8: 15 Minute Lightning Plot of east values reached near 700 mb with the 3 central Washington 0530 UTC 21 August g/kg values near 600 mb with lower 2009 values to the east and west. Higher RH was also present near the base of the While the upper level forecasting potentially unstable layer, which would diagnostics were opposed, the measures be released with the forcing from the used to assess instability, favored strong short wave trough. convection. ULLR at 06 UTC (Figure 11)

Figures 9&10: 06 UTC 21 August 2009 1.5 PV surface showing pressure of dynamic tropopause (black contours) and wind barbs (brown). 06 UTC 22 August 2009 250mb Geopotential heights (black contours) and wind barbs (black with color gradient for speed)

Figures 11&12: 06 UTC 21 August 2009 Upper Level Lapse Rate Contours. 06 UTC 21 August 2009 Θe (Kelvin, blue contours), mixing ratio (g/kg, colored filled with scale on left), and wind barbs (black) NARR cross‐section. Approximate location from west of Portland, OR to north of Coeur d'Alene, ID.

4. Conclusion

Monsoonal moisture along with for missing or busting a forecast remains Pacific trough dynamics represents a high. common occurrence for dry lightning in Isentropic analysis represents the western United States. This procedure most comprehensive tool within the allows forecasters to diagnose lightning procedure while the ULLR might be the busts ahead of time in the absence of most important. Looking at other events strong synoptic or traditional evidence of that did or more importantly did not thunderstorm development. Dry materialize, the ULLR threshold of 7.5 thunderstorm development usually C/km holds true. This may be due it occurs at the edges of the best dynamics encompassing the other three tools and moisture, which makes it somewhat effects into one value. However, this does of a difficult forecast. Using traditional not mean that this should be the only step methods and indices can leave a one should consider. It does not resolve forecaster unsure of the potential for the issue of moisture in the atmosphere, thunderstorm development. And with which remains equally important. Jet many of these events having narrow streak and dynamic tropopause analyses gradients of lightning and/or with HLTT provide support to the top two thunderstorm development, the potential steps and more confidence in the forecast of dry thunderstorms. The procedural Some limitations and possible works best as sum of all its parts and not additions to this procedure could better individually. The collocation of each step distinguish the relationship between along with the magnitude of each storm speed, motion and precipitable determine if an event will happen and if it water. This would better exhibit the does, the scale of it. potential for dry versus wet When there are less significant thunderstorms. Storm movement and features of each procedural step, the development along with the amount of chance of not having dry thunderstorms precipitable water would greatly impact does not necessarily go down, but the how much rainfall occurs over an area. magnitude of the event will be less. This affects the potential for ignition and Smaller events can be fueled by local or could negate the effects of lightning mesoscale processes, which are not strikes collocated with the rainfall. discussed in this procedure. However, the Additionally, surface RH and fuel focus of this procedure is to detail a conditions could be added to provide a forecast for dry thunderstorms especially more comprehensive dry thunderstorm events that will stress local fire initial forecast tailored to fire managers, who attack capabilities. feel the most impact of these events.

Acknowledgements. Dr. Mike Kaplan, Hauss Reinbold, Bo Tan, and NWS WFO Reno.

REFERENCES

Crimmins, Michael A., 2006. Synoptic Climatology of Extreme Fire‐Weather Conditions Across the Southwest United States. International Journal of Climatology, 26, 1001‐1016.

Hirschberg, P. and J. M. Fristch, 1991. Tropopause Undulations and the Development of Extratropical . Part I: Overview and Observations of a Event. MWR, 119, 496‐517.

Hirschberg, P. and J. M. Fristch, 1991. Tropopause Undulations and the Development of Extratropical Cyclones. Part II: Diagnostic Analysis and Conceptual Model. MWR, 119, 518‐ 550.

Milne, R. M., 2004. A Modified Total Totals Index for Thunderstorm Potential over the Intermountain West. WR Technical Attachment No. 04­04.

Rorig, Miriam L. and Sue Ferguson, 1999. Characteristics of Lightning and Wildland Fire Ignition in the Pacific Northwest. Journal of Applied , 38, 1565‐1575.

Rorg, Miriam L. and Sue A. Ferguson. The 2000 Fire : Lightning‐Caused Fires. Journal of Applied Meteorology, 41, 786‐791.

Rorig, Miriam L., Mckay, Steven L., Ferguson, Sue A., and Paul Werth. Model‐Generated Predictions of Dry Thunderstorm Potential. Journal of Applied Meteorology and Climatology, 46, 605‐614.

Wallmann, J., 2004. A Procedure for Forecasting Dry Thunderstorms in the Great Basin using the Dynamic Tropopause and Alternate Tools for Assessing Instability. WR Technical Attachment No. 04­04. August 3, 2004.