2008 Forest fires in the Northern California, USA

Abstract

In recent years, severe wildland fires have occurred worldwide in regions under the influence of a Mediterranean climate. Recent fires in Greece (August 2007), southern California (October 2007), northern California (June 2008), and Victoria, Australia (February 2008) resulted in human fatalities. This paper describes one of the largest outbreaks of forest fires in northern California that occurred during June of 2008. This widespread outbreak of over 800 fires was caused by dry lightning storms that followed one of the driest springs on record. Remotely sensed data from NASA (TERRA, GOES, and MODIS) and weather data from NOAA (vertical air temperature profiles) are used to describe and explain the large fire outbreak. Results of field research were also discussed to explain the reasons for the long lasted active forest fires after the "historical" lightning ignition.

Full text

The Eighth Symposium on Fire and Forest Meteorology, 13-15 October 2009, Kalispell, Montana
2.5 2008 FOREST FIRES IN THE NORTHERN CALIFORNIA, USA
Hiroshi Hayasaka* and Carl N. Skinner**
* Hokkaido University, Sapporo, Hokkaido, Japan
** U.S. Forest Service, Pacific Southwest Research Station, Redding, California
1. INTRODUCTION
An unusually large outbreak of fires for northern
California occurred in June 2008. Unprecedented dry
thunderstorms with severe lightning occurred under
very dry conditions due to the rainless weather since
March. More than 8,000 lightning flashes were
observed mainly in northern California and western
Nevada. After the historical dry thunderstorms, more
than 800 forest and wildland fires started and many
lasted more than a month under dry and hot summer
conditions. The peak number of fires totaled 2,696 and
total burned area reached about 4,580 km2. 511
houses were burned. Unfortunately, 15 firefighters lost
their lives (2008 Lightning Siege Overview,
International CNN.com).
One of the authors from Japan could have a chance
to survey forest fire situation in northern California and
visited the U.S. Forest Service’s Pacific Southwest
Research Station, Redding, California. Survey period
was the week of July 9, 2008 that coincided with the
largest peak in burned area. In this preliminary report,
lightning data of these historical dry thunderstorms,
weather data measured at Redding, fire (hotspot) data,
weather maps, emagrams, and various satellite images
were analyzed to explain occurrence of the severe
lightning and the following fires.
There are two reports at least about this event as far
as the authors knew. One is “Sacramento Fire Weather
annual summary- 2008” and another one is a draft
paper written by M. Burger. Two papers mainly
described mechanism of historical dry thunderstorm
only from meteorological point of view.
2. SURVEY OVERVIEW
Fig.1 is a map of northern California and a part of
Nevada showing the distribution of lightning (dots) and
fire (gray rectangles) is shown in Fig.1 Our focused
research areas were shown in this Fig.1 by using
rectangles numbered from 1 to 5. Both authors visited
the Motion fire (area 1) and the lead author visited
areas 2 to 5 alone.
Picture 1 showed the situation of “Motion” fire on
July 11, 2008. This picture was taken from the center of
Redding looking west. “Motion” fire was very active
quickly covering a large area so it appeared that each
mountain had its own fires. So many fires in adjacent
mountains imply both dense lightning and rapid fire
expansion by spot fire.
Fig.1 Distribution of Lightning and Fires (Hotspots)
Picture.1 “Motion” Fire Near Redding
(July 11, 2008)
3. BACKGROUND OF LARGE FOREST FIRES DUE
TO DRY THUNDERSTORM
3.1 Special features of climate and region in
northern California
Northern California has a pronounced
Mediterranean type climate influenced by strong high
pressure in the adjacent Pacific that blocks frontal
systems from northern California in summer. The
center of high pressure is usually located at the
neighborhood of about 50 degrees north latitude about
and 150 degrees west longitude. This contributes to a
prevailing northerly wind in the area along the
Californian seashore. These winds cause upwelling off
the California coast. Due to upwelling, sea surface
temperatures near Californian are kept low even in mid
summer. The center of upwelling is located offshore
from San Francisco to Mendocino. Thus, mean sea
surface temperature in August is only 11.1 oC.
This low summer sea temperature helps create
stable atmospheric conditions under 300m height.
Summer fog near the California coast is often a result
of the cool California Current offshore. Lightning is not
* Corresponding author address: Hiroshi Hayasaka,
Graduate School of Engng., Hokkaido Univ., Sapporo,
060-8628, JAPAN; e-mail: hhaya@eng.hokudai.ac.jp

The Eighth Symposium on Fire and Forest Meteorology, 13-15 October 2009, Kalispell, Montana
common under such atmospheric stability or
subsidence inversion.
On the other hand, weather in inland California is
greatly different from the coast. Fog and cool wet air of
the Pacific Ocean do not move inland more than a few
km. Instead, the inland area of northern California is
usually hot and dry in summer. As a result, a big
temperature difference exits between foggy coast with
air temperature of 10 oC and inland basin with 40 oC in
summer.
In addition to the above-mentioned climatic
features, northern California occasionally experiences
tropical maritime air masses. These later conditions are
often referred to as “pineapple express” and provide
northern California with large amounts of rain such as
in the winter of 1997. These events clearly showed that
the largest amounts of rain occurred where wet air
flowing from the Pacific Ocean hit the western slopes of
northern California mountains.
Thus, we suggest that the dry thunderstorms (Rorig,
M.L. et al. 2006) of June 2008 may have been the
result of tropical maritime air masses from near Hawaii
islands. In other words, if wind direction changed from
northerly to westerly or southerly, then the influx of
moist, unstable air would help create conditions for
thunderstorms because westerly or southerly winds are
usually moist.
3.2 Recent Precipitation Trend
Mean and monthly precipitation in 2008 in Redding
(California Weather Data) are shown in Fig.2. Annual
mean precipitation in Redding is 999mm. But annual
mean precipitation in 2007 and 2008 was 551 and
550mm respectively. This limited precipitation condition
started from around May 2006 and lasted until May
2009.
Fig.2 Precipitation in Redding
2008 monthly precipitation in Redding is shown in
Fig.2 along with mean monthly precipitation. Bars with
solid color are for 2008. From Fig.2, you will find
rainless months started in March. Actually, the last
rainfall of 8 mm was observed on 24 February 2008.
After that, no significant amount of rain was observed
until the end of September.
4. CONDITIONS FOR THE OCURRENCE OF THESE
DRY THUNDER STORMS
4.1 Time series of lighting occurrence
Lightning position on June 20 and 21, 2008 are
shown with small dots in Fig.1. Lightning started over
the sea on the evening of June 20. Lightning position
moved from the southwest to the northeast. Much
lightning, more than 8,000 flashes, occurred on June
21. This thunderstorm is called “historical” dry
thunderstorm because the great number of lightning
flashes and damages due to forest fires were unusual.
Fig.3 shows hourly lightning strikes from 16:00 of
June 20 to 14:00 of June 21. Fig.3 covers only a part of
the northern California area (North latitude: 39-42o,
West longitude: 122.5-125o) for discussion of forest
and wildland fires in the northern California. As a result,
total number of lightning flashes (2,334) is less than
previously stated. Fig.3 clearly shows lightning
occurrence peaked at 7:00 of June 21. This implied
lightning was not a so-called thermal lightning that tend
to peak in the middle of the day to early evening.
Hour
Number of Lightning Flashes
100
200
300
400
0
Total Number of Lightning Flashe s: 2,334
Detection Area:
Northwest California
39.0-42.0 N, 122.5-125.0 W
June 20-21, 2008
16 18 20 22 10 12 140 2 4 6 8
Day Number
172 173
Fig.3 Lightning occurrence tendency in northern
California
4.2 Weather Map, Clouds and Jet Stream
Surface weather map (NCEP/NCAR Reanalysis 1)
at 12:00 on June 21, 2008 is shown in Fig.4. From
Fig.4, two high-pressure zones were found north of
Hawaii and in the Colorado Plateau. Four low pressure
zones located south of Alaska Peninsular, in the east of
Gulf of Alaska, west of California, and in northern Baja
California.
Fig.4 Surface Weather Map at 12 UTC June 21

The Eighth Symposium on Fire and Forest Meteorology, 13-15 October 2009, Kalispell, Montana
Satellite image take by GOES (California Regional
Weather Server, San Francisco State University) at
18:00 UTC on June 21 is shown in Fig.5. Four vertical
and horizontal dotted lines in Fig. 5 show west
longitude of 120 and 135 degrees, and north latitude of
45 and 60 degrees respectively. Vortex clouds named
C1 and C2 are seen in Fig. 5 and they were formed
corresponding to low-pressure zones in Fig.4. The size
of the vortex clouds of C2 is about 1,400 km from
around 140 degrees west longitude to west coast of
United States and about 2,000 km from about north
latitude of 30 to 50 degrees. This size is not so small as
middle-latitude storm.
In addition to C1 and C2, C3 was found in northern
California. C3 clouds made historical dry
thunderstorms and C3 appears to have been
independent from C2 clouds. According to GFS Model
analysis, the jet stream was approaching from the
northwest coast of United States. C3 clouds may have
formed with the help of tropical moisture flows from
near Hawaii and near the Gulf of California. When they
entered into northern California, the jet steam may
have helped the formation of C3 clouds.
Fig.5 Surface Weather Map at 12 UTC June 21
4.3 Emagrams
To study the mechanism of these historical dry
thunderstorms, emagrams (Atmospheric Science,
University of Wyoming) were made. Fig.6, shows
emagrams from the Oakland Airport near San
Francisco.
Fig.6 Emagrams in Oakland
Emagram at 00:00 on June 21 is shown in the left
panel in Fig.6. The right side panel in Fig.6 is for 12:00
on June 21. It is apparent there was a big difference
between their air temperatures up to 800m. Air
temperature at height 6m from ground changed from
34.2 oC (relative humidity, RH, 16%) to 22.2 oC (RH,
41%). From right-hand side figure of Fig. 6, large
environmental lapse rate (15.5~18.2 oC/1000m) was
observed up to a height of about 400m.
Table 1 shows more detail in temperature
conditions up to a height of 2,000m at 12:00 UTC on
June 21 in 2008. Table 1 also shows temperatures in
recent yeas from 2004 to 2007 for comparison. From
Table 1, you will see that temperatures in 2008 at any
height were higher than those of other recent years.
Especially, the temperature of 29.1 oC at a height of
500m was very high. There was also a large lapse rate
of 7.4 oC/1,000m(ΔT2=T2000-T500). This lapse rate
was larger than in the other four years. Thus, the
atmosphere in 2008 was “conditionally unstable”.
Table 1. Comparison of Temperatures (12:00 UTC
on June 21) at Various Heights in Recent Years
Height
(m)
2004
2005
2006
2007
Mean
’04-‘07
2008
ΔT1
6
13.0
13.4
14.0
12.4
13.2
22.2
9.0
500
9.2
10.3
20.9
9.1
12.4
29.1
16.7
1,000
20.5
10.2
25.1
19.9
18.9
27.0
8.1
2,000
16.0
9.9
17.3
14.6
14.5
18.0
3.6
ΔT2
6.8
-0.4
-3.6
5.5
2.1
-11.1
-13.2
4.4 Observation Clouds Using Satellite Images
A satellite image take by Terra at about 10:00 LST
on June 21 was shown In Fig.7.
Fig.7 Terra Imagery at about 10:00 LST on June 21
Redding is located near the top center of Fig.7. In
this image Redding was under relatively large white
cloud that was part of the historical thunderstorms. The
inland clouds are considerably different in shape
compared to the clouds over the Pacific Ocean. Many
of the inland clouds tend to have an origin. Arrows in
Fig.7 show such origins. These clouds may form due to
such as islands at sea, a cape near the coast, or hill
and mountain on land. Such prongs will create
atmospheric disturbances. Three clouds in Fig.7 shown
with arrows may have been formed by topographical
prongs, but the other two clouds with arrows may have
been created by the effect of a forest fire. Actually,
Terra MODIS detected hotspot (forest fire) on June 20
near right-pointing arrow in Fig.7.

The Eighth Symposium on Fire and Forest Meteorology, 13-15 October 2009, Kalispell, Montana
5. ANALYSIS OF FIRE AND WEATHER CONDITION
5.1 Forest Fires (Hotspots) and Fire Occurrence
Rate
In Fig.1, fire locations were shown with gray
rectangles and lightning locations were shown with
dots. Both symbols exaggerated size in the image.
Also hotspots were detected two days after the
historical lightning event. This is due to MODIS not
being able to detect a forest fire until it could become
relatively large.
To calculate the fire occurrence rate, hotspot will
not be suitable data because MODIS could not detect a
small forest fire. For this, two figures from fire fighter
teams and news are used. The total number of
lightning flashes was 5,064 for the region of northern
California (North Latitude: 37.5-42 degrees, West
Longitude: 120-124.4 degrees). Formula for fire
occurrence rate is show below.
FOR (Fire occurrence rate) = (Number of fires) /
(Number of lightning flashes) x 100 (%)
Fire occurrence rate is also called the “ignition rate”.
Here I used fire occurrence rate.
The number of fires was 2,096 according to fire
fighter teams and 800 from news. Their fire occurrence
rates were 41.4 % and 10% respectively. There is a big
difference. This is simply because fire fighter teams
could include small fires and new fires caused by
spotting fire. But we should pay more attention on high
fire occurrence rate. Some explanation is needed for
future fire fighting.
5.2 Time Series of Forest Fires (Hotspots)
Fig.8 shows the number of hotspots become 731
just two days after ignition. This number is similar to the
fire number from news. This is the number of fire areas
that grew large enough to be detected.
0
200
400
600
800
1000
170 180 190 200 210 220
Day Number
230
Number of Hotspots
June July August
Lightning
25
30
35
40
Hotspot in Northern Clifornia
(N 38-42, W 120-125)
Hotspot Max. Air Temp. at Redding
2008
Temperature C
O
Fig.8 Fire (hotspot) and Air Temperature after
Ignition (North Latitude: 38-42 degrees, West
Longitude: 120-125 degrees)
After this first hotspot peak in Fig.8, the number of
hotspots decreased. But a second peak started in early
July when maximum air temperature reached around
40 oC. Maximum hotspot number occurred in this
second peak and it was 955. After the second peak,
the number of hotspots decreased but more than 200
hotspots were observed until the top of August.
5.3 Lightning and Forest Fires near Redding
To learn more about the relationship between the
historical thunderstorms and the resulting
lightning-caused fires, the distribution of lightning and
fire locations were plotted in Fig.9. Here, the total
number of lightning flashes was 426. Lightning
locations are shown by cross marks with three different
colors. Different colors were used to distinguish
lightning occurrence time.
Highest number of lightning flashes was 82
observed at 11:00 LST on June 21. Three dense
lightning groups were found in Fig.9. They were named
1, 2, and 3. They were highlighted by ovals. Direction
of each group was from the southwest to the northeast.
This direction was same as in Fig.1. Three groups of
lightning follow the shape of the land, in other words,
along or above mountains.
Lightning began around 5:00 LST on June 21. Until
11:00 LST, 245 lightning flashes were observed mainly
in zones of 1 and 2 in Fig.9. From 11:00 to 14:00 LST,
the number of lightning flashes was 160 and they hit
mainly in zone 3.
Further investigation is needed to explain fires
occurrence in zone 2 because there was a big
difference in the fire (hotspot) occurrence rate among
the three zones. Understanding the reason for many
hotspots in zone 2 would be helpful for future fire
fighting.
Fig.9 Distribution of Lightning and Fire Near
Redding
6. CONCLUSIONS
1. Clouds of the historical thunderstorm were identified
on various satellite images such as Terra and GOES.
The mechanism leading to formation of the historical
thunderstorm was not so clear. This may suggest that
there was complex cause. A detailed study of the
historical thunderstorm should be undertaken.

The Eighth Symposium on Fire and Forest Meteorology, 13-15 October 2009, Kalispell, Montana
2. High fire occurrence rate of around 10% (more than
800 forest and wildland fires ignited by 8,000 lightning
flashes) is likely due to highly dried vegetation under
strong drought condition and the lack of significant
moisture with these storms.
3. The activity of these fires appeared to be associated
with air temperature and wind speed and direction.
Highest peak of hotspot (fire) was observed when air
temperature exceeded 40 oC.
REFERENCES
M.L. RORIG, S.J. MCKAY, S.A. FERGUSON and P.
Werth, 2006 : “Model-Generated Predictions of Dry
Thunderstorm Potential”, J. Applied Meteorology and
Climatology, Vol. 46, pp. 605-614. and etc.
2008 Lightning Siege Overview, California Department
of Forestry and Fire Protection, http://www.fire.ca.gov
/index_incidents_overview.php
International CNN.com/US, Hundreds of post-lightning
fires burn in California - CNN.com, 08/06/25 9:06. and
etc.
California Weather Data, Agriculture and Natural
Resources, University of California. http://169.237.
140.1/WEATHER/ wxretrieve.html
NCEP/NCAR Reanalysis 1: Surface, Earth System
Research Laboratory, Physical Sciences Division,
NOAA Research, U.S. Department of Commerce.
http://www.cdc.noaa.gov/data/gridded/data.ncep.
reanalysis.surface.html
Satellite images, California Regional Weather Server,
San Francisco State University. http://squall.sfsu.edu/
Department of Atmospheric Science, Collage of
Engineering, University of Wyoming. http://weather.
uwyo.edu/upperair/sounding.html?region=seasia
Unknown Author, Sacramento Fire Weather annual
summary - 2008, http://www.wrh.noaa.gov/sto/fire_wx
_verification_2008.pdf
M. Burger : “The northern California lightning event of
June 20-21, 2008”, 23rd Conference on Weather
Analysis and Forecasting/19th Conference on
Numerical Weather Prediction, June, 2009.
P. J. Bartlein, S. W. Hostetler, S. L. Shafer, J. O.
Holman and A. M. Solomon, 2008, Temporal and
spatial structure in a daily wildfire-start data set from
the western United States (1986–96), Int. J. of Wildland
Fire, 17, 8–17. .