Santa Anna and Anomalies to forecast them By Richard H. Grumm National Weather Service Office State College, PA

1. INTRODUCTION pressure builds over the Great Basin low pressure develops offshore. Many During the latter part of October 2007, studies have been conducted related to strong Santa Anna winds and dry the meteorology of Santa Ana winds conditions produced devastating fires in events. Raphael (2003) provides and Southern . News reports of excellent summary of these events. rapidly moving fires appeared by 20 Small (1995) provides a summary of a October 2007 and the destructive forces 1993 event which was also associated of these fires dominated news reports with brush fires. from 21-27 October 2007. Weaker winds and more humid conditions diminished Rapheal (2003) developed a 32-year the fire activity after the 26th. dataset to quantify Santa Ana events. He noted that there are approximately 20 Moderate Resolution Imaging events per season and most events lasted Spectroradiometer (MODIS) imagery about 1.5 days. The most active months captures the fires and NASA began a were November through January (see his case page with images from 21-29 Figure 3) though the season began in October capturing the event. Figure 1 September and lingered into April. Key show MODIS images from the NASA features appeared to be low dew points on the 23 and 24th of October 2007. The and a strong pressure gradient from the large plumes of smoke directed from Great Basin to coastal California. east to west characterize the strong Santa Ana winds which are common in many The key conditions with Santa Ana fire cases1 (Raphael 2003). The winds, events have not been associated with when coincidental with the fires drive anomalies in mean-sea level pressure them westward. fields. One goal of this study is to show the pressure anomalies over the Great The Santa Ana is a hot dry easterly Basin and during or northeasterly wind that blows from known Santa Ana events and those the deserts of Nevada and southern events which were associated with fires California into the eastern Pacific when such data was available. The (Glickman 2000). The wind is also anomalies were computed as outlined in characterized as foehn-type event named Hart and Grumm (2001). after the Santa Ana river valley and the pass of the same name. The stronger The goal of this paper is to present the events normally occur in the mid- conditions associated with the Autumn to early spring when strong high devastating fires and the persistent Santa Ana wind event of 21-29 October 2007.

1 Many fires are of human origin and one of the Forecasts and analysis of anomaly data larger fires of October 2007 was likely set by a will be presented. These data will be young man. contrast to other known and documented Cases were obtained by searching the significant Santa Ana events. web and articles published on Fires and Santa Ana winds. This was neither a 2. METHODS AND DATA comprehensive or exhaustive study. The 6 January 2002 event was found on the All model data were retrieved in near- NASA MODIS site. Imagery of dust and real-time from NCEP. The data shown a few fires were available along with a here are primarily from the NCEP brief description of the event. The Global Ensemble Forecast System Shekell, Fire in California on 3 (GEFS). The data are shown with December 2006 was a massive fire in respect to the 30-year climatology. The southern California which was captured forecasts are subtracted from the 30-year on MODIS imagery. The fire of late mean and normalized by the value of 1 October 2003 was another large fire standard deviation on the particular day captured by MODIS imagery. Imagery and hour. Thus all data are shown as of the fire was archived by NASA on the standardized anomalies. 24-30 October 2003 for this massive event which had similar east to west Re-analysis data is used to show smoke plumes. Figure 5 shows this conditions of similar events. All re- image as a case study presented in this analysis data is displayed showing the document. field and its associated standardized anomaly (Grumm and Hart 2001). Table 1 shows Santa Ana winds recorded at the National Weather All re-analysis, model, and anomaly data Service office in , California. were displayed using GrADS. Not all events are recorded. Several of these cases were reviewed for this study. Figure 2 Conditions over the southwestern United States from the NCEP/NCAR global re- analysis data valid at 0000 UTC 25 October 2003. Data shown include a) mean sea level pressure (hPa) and anomalies, b) precipitable (mm) water and anomalies, c) 850 hPa winds and u-wind anomalies, and d) 850 hPa winds and v-wind anomalies. Anomalies are computed using the value minus the mean for that date divided by 1σ for that date.

Figure 2 As in Figure 2 except valid at 0000 UTC 26 October 2003.

In

Figure 3 As in Figure 2 except valid at 0000 UTC 27 October 2003.

3. RESULTS addition to the strong low-level , the event of October 2003 i. October 2003 event was associated with a massive upper- level ridge as shown in Figure 4. At The long-lived Santa Ana event of 0000 UTC 27 October the 500 hPa and October 2003 produced massive fires 700 hPa heights were +2SDs above and shared many of the characteristics normal over western North America and associated with the October 2007 event. the eastern Pacific. The low-level 925 Figure 2 shows the large scale conditions hPa winds were anomalously strong over the southwestern United States at from the northeast. 0000 UTC 25 October 2003. Key features include the above normal mean- An examination of other documented sea level pressure over the Great Basin Santa Ana events suggests that the (Fig. 2a), the below normal moisture in anomalous ridge is often associated with the precipitable water field (Fig. 2b), and stronger Santa Ana events. A thorough the strong northeast winds south of the study has not been accomplished to anticyclone (Figs. 2c&2d). The prove the significance of these features. anomalous easterly winds pushed The large ridge likely contributes to southward on the 26th (Fig. 3). larger scale subsidence and drying. The subsidence may play a role in the It should be noted that the anticyclone development of an inversion near the over the Great Basin is similar to the ridge tops. Other cases examined (not pattern presented by Rapheal (2003). shown) showed similar patterns with Figure 5 MODIS image of California fires 26 October 2003. anomalous from the surface diminish slowly to about +1SD above into the mid-troposphere. normal by the 24th. ii. October 2007 event In addition to the strong MSLP anomalies, the NCEP GEFS forecast Forecasts of the surface pressure over significantly above normal easterly the United States from Forecasts winds over the region. Note the -4SD initialized at 0000 UTC 20 October 2007 wind anomalies at 850 hPa at 1800 UTC valid at 1800 UTC 21 & 22 October 22 October 2007. Anomalous easterlies 2007 are shown in Figure 6. were forecast to persist through 0000 UTC 25 October 2007. The key features in these forecasts was the emergence of above normal high A sequence of GEFS forecasts of 500 pressure over the Great Basin on the 21st hPa heights from 0000 UTC 19 October then the appearance of +3 to +4SD through 0000 UTC 21 October 2007 is pressure anomalies over Nevada and shown in Figure 8. These forecasts are California on the 22nd. In addition to the all valid at 0000 UTC 23 October 2007. strong anomalies, there was good The key feature shown is the anomalous agreement between ensemble members 500 hPa ridge to the west of the affected as to the location and magnitude of the area. The anomalous surface fields (not large anticyclone. Though not shown, shown) were east of the upper-level the MSLP anomalies were forecast to system over the Great Basin. The data show the ability of the GEFS to forecast Figure 6. NCEP GEFS forecasts of mean sea-level pressure (hPa) from forecasts initialized at 0000 UTC 20 October 2007 showing left) forecasts valid at 1800 UTC 21 and right) valid at 1800 UTC 22 October 2007. Upper panels shows each members 1008 and 1020 contours with spread and lower panels show the ensemble mean forecasts and the departures of these fields from the 30-year climatology in standard deviations from normal.

the key features associated with these Southern California. News reports of events. rapidly moving fires appeared by 20 October 2007 and the destructive forces Figure 9 shows successive forecasts of of these fires dominated news from 21- the 850 hPa winds. Upper panels show 27 October 2007. Weaker winds and the u-wind anomalies. These forecast all more humid conditions diminished the showed abnormally strong u-winds and fire activity after the 26th. large u-wind anomalies over the region affected by the Santa Ana winds. These An attempt was made here to show some data imply the value of wind anomalies of the key synoptic scale features as in aid in forecasting these events. associated with Santa Ana wind events in southern California. The fires, when 4. CONCLUSIONS present or ignited during such an event can spread rapidly due to the During the latter part of October 2007, combination of dry conditions and strong Santa Anna winds and dry strong warm winds. No attempt is made conditions produced devastating fires in here to predict fires. Forecasting fires and the threat of easily ignited fires likely requires dry conditions at low-levels as indicated by low level dew points and dew point depressions, the lack of any recent precipitation2, dry soils and dry fuels. Cloud cover and stability can also impact the fire threat. An unstable and deep boundary layer may favor feeding oxygen rich air into the fire. Many of these parameters can readily be extracted from model grids. The intangible is the human factor. Models may produce very light convective rain indicating the potential for lightning which could ignite a natural fire.

The focus here is on the more tangible Santa Ana winds. The signal for these wind events is more readily identifiable than those Figure 7. As in Figure 4 except GEFS 850 hPa winds and anomalies which may contribute to the valid at 1800 UTC 21 October showing a) 850 hPa winds and u-wind anomalies and b) 850 hPa winds and v-wind anomalies. threat of fires. The data shown here suggest that the use of climate anomalies may aid in diagnosing and patterns Fires in many parts of the United States associated with the Santa Ana winds. are initiated by lightning and humans. The GEFS data presented here indicated During dry conditions and large that the GEFS forecast the pattern quite subsidence event; such as is present in well. Additionally, the anomalies of key Santa Ana winds; can rapidly fan and fields, indicating a significant event were produce fast moving fires. It should be well forecast. These features included: noted that many fires are likely of • Anomalous mean sea- human origin unless they were ignited level pressure over the by early convection prior to the onset of Great Basin. the winds. Another potential source of • Anomalous 500 hPa fires could be due to blown down heights over the region electrical lines.

2 This list was slight refined based on conversations with Jeremy Ross of Zedex. Figure 8 GEFS forecasts of 500 hPa heights valid at 0000 UTC 23 October 2007 from forecasts initialized (left to right) at 0000 UTC 19 October, 0000 UTC 20 October and 0000 UTC 21 October 2007. Upper panels show the 500 hPa spaghetti plots with the 5460 m and 5760 m contours by each member. Shading shows the spread. Lower panels show the ensemble mean and the standardized anomalies. • And anomalous 850 hPa associated with fires (Small 1995). The u-winds over the region. features with associated with this event are shown in Figures 10 & 11. The Forecasting strong events surface fields show the classic pattern require may require3: with a massive anticyclone over the • Wind 20 to 30 knots of wind Great Basin. The anticyclone central orthogonal to the mountain range; pressure was 1040 hPa with an • Many events are associated with expansive area of +2SD to +3SD above subsidence in the mid-levels of the normal pressure anomalies. Strong 850 atmosphere often characterized by hPa winds were present over southern negative vorticity advection. California, implying strong winds nearly • Channeling of winds down a valley orthogonal to the mountains. The PW in the general direction of the wind. fields showed below normal moisture in • Ascent in low-levels beneath the the source region. subsidence may create or enhance the inversion over the mountain The upper-level pattern showed a strong range ridge at both 500 and 700 hPa. The • The formation of mountain waves upper-level height anomalies were on which amplify with the inversion. the order of +2 to +3 SDs above normal. The strong ridge and its orientation In addition to the 2003 and 2007 cases, implied strong subsidence over the the event of 26-27 October 1993 was region, a key component to many Foehn also a significant Santa Ana event wind events. Similar to the 850 hPa winds the 925 winds were also shown to be nearly orthogonal to the mountains 3 List is based on experience and conversations and around 1 to 2SDs above normal. with J. Michael Fritsch and Richard James. Figure 9. As in Figure 8 except showing 850 hPa winds with u-wind anomalies in upper panels and v-wind anomalies in the lower panels. investigate the details meso-meteorology Figure 12 shows the conditions at 1800 of these local wind events. UTC on 26 October 1993. These data use the NARR data and thus are of The 3 cases presented suggest that significantly higher resolution than the climatic anomalies could be used to GR data. In fact, several wind vectors identify strong foehn (Fohn) wind had to be skipped in order to see the events. The anomalies alone are not anomalies. These 32km data showed sufficient to forecast these events, considerably higher winds and though they can point to significant significantly strong anomalies in the features associated with them. Other winds. Other fields, such as the MSLP fields need to be examined, such as and 500 hPa heights were more similar vertical motions near the mountains and to those shown by the GR data. The stability. The presence of an inversion NARR data also indicated subsidence just above the mountain tops can help between 850 and 500 hPa (not shown) amplify mountain waves and thus another feature often associated with increase the wind speed of the winds foehn wind events. descending the mountain slopes.

Though not shown, the 1.25x1.25 5. Acknowledgements: degree, JMA data showed similar patterns and features. It was more Mike Fritsch, Richard James, and similar to the coarser GR data and did Jeremy Ross for forecast information not have the stronger low-level winds and materials related to forecasting indicated in the NARR data. The NARR downslope wind events and fires. data is a potential method to further 6. References: Figure 10. As in Figure 2 except valid at 1200 UTC 26 October 1993.

Figure 11 As in Figure 3 except valid at 1200 UTC 26 October 1993.

Chelton, D.B., M.H. Freilich, and S.K. Esbensen, Satellite observations of the Rev.,129,2426-2442. wind jets off the Pacific coast ofcentral America. Part I: Case studies and statistical characteristics. Mon. Wea. Lynn, R.J., J. Svejkovsky, Remotely Rev., 128, 1993-2018,2000. sensed sea surface temperature variability off California during a "Santa Glickman, T., 2000: Glossary of Ana” clearing, J. Geophys. Res., 89, Meteorology, 2d. Amer. Meteor. Soc. 8151-8162, 1984. Boston, MA. Moulin, C, H.R. Gordon, V.F. Banzon, Grumm,R.H. and R. Hart, 2001: R.H. Evans, Assessment of Saharan dust Standardized Anom alies Applied to absorption in the visible from SeaWiFS Significant Cold Season Weather imagery, J. Geophys. Res., 106, 18239- Events: Preliminary Findings. Wea. and 18249, 2001 Forecasting,16,736-754. Raphael, M.N. 2003: The Santa Ana Hart, R.E. and R.H. Grumm, 2001: winds of California. Earth Interactions. Using Normalized Climatological 7,1-13. Anomalies to Rank Synoptic-Scale Events Objectively. Mon. Wea. Small, I., Santa Ana winds and the fire

Figure 12 NARR data compared to GR anomalies showing conditions at 1800 UTC 26 October 1993. Wind vectors are scaled to skip every 2 grid points for clarity. outbreak of fall 1993, NOAA Technical Memorandum NWS WR-230 June 1995.

Sommers, W.T., LFM forecast variables related to Santa Ana wind occurrences, Mon. Wea. Rev., 106, 1307-1316, 1978. Santa Ana Events Date Event Comments Type 21-22 Nov SAW 1956 5-6 Nov SAW 1961 26 Sep 1963 SAW

23 Nov 1986 SAW 20 Jan 1987 SAW 6-7 Feb 1987 SAW 18 Nov 1987 SAW 12-13 Dec SAW 1987 16-19 Feb SAW 1988 30 Nov 1988 SAW 8 Dec 1988 SAW 28 Nov 1989 SAW 11 Dec 1989 SAW 26-27 Oct SAW Shown here 1993 Small (1995) study 25-27 Oct SAW Massive 2003 wild fires 200 killed. Used as reference case here. 6-7 Feb 2006 SAW Ensemble forecasts shown here Major fires Table 1 San Diego WFO wind events deverived from their office weather history site. SAW means Santa Ana Wind event. Return to text.