ArticlePDF Available

California heat waves in the present and future

Authors:

Abstract and Figures

Current and projected heat waves are examined over California and its sub-regions in observations and downscaled global climate model (GCM) simulations. California heat wave activity falls into two distinct types: (1) typically dry daytime heat waves and (2) humid nighttime-accentuated events (Type I and Type II, respectively). The four GCMs considered project Type II heat waves to intensify more with climate change than the historically characteristic Type I events, although both types are projected to increase. This trend is already clearly observed and simulated to various degrees over all sub-regions of California. Part of the intensification in heat wave activity is due directly to mean warming. However, when one considers non-stationarity in daily temperature variance, desert heat waves are expected to become progressively and relatively less intense while coastal heat waves are projected to intensify even relative to the background warming. This result generally holds for both types of heat waves across models. Given the high coastal population density and low acclimatization to heat, especially humid heat, this trend bodes ill for coastal communities, jeopardizing public health and stressing energy resources.
Content may be subject to copyright.
California heat waves in the present and future
Alexander Gershunov
1
and Kristen Guirguis
1
Received 6 July 2012; revised 10 August 2012; accepted 20 August 2012; published 27 September 2012.
[1] Current and projected heat waves are examined over
California and its sub-regions in observations and downscaled
global climate model (GCM) simulations. California heat wave
activity falls into two distinct types: (1) typically dry daytime
heat waves and (2) humid nighttime-accentuated events (Type I
and Type II, respectively). The four GCMs considered project
Type II heat waves to intensify more with climate change than
the historically characteristic Type I events, although both types
are projected to increase. This trend is already clearly observed
and simulated to various degrees over all sub-regions of
California. Part of the intensification in heat wave activity is
due directly to mean warming. However, when one considers
non-stationarity in daily temperature variance, desert heat
waves are expected to become progressively and relatively less
intense while coastal heat waves are projected to intensify even
relative to the background warming. This result generally
holds for both types of heat waves across models. Given the
high coastal population density and low acclimatization to
heat, especially humid heat, this trend bodes ill for coastal
communities, jeopardizing public health and stressing energy
resources.
Citation: Gershunov, A., and K. Guirguis (2012),
California heat waves in the present and future, Geophys. Res. Lett.,
39, L18710, doi:10.1029/2012GL052979.
1. Introduction
[2] The flavor of California heat waves is changing: they are
becoming more humid and therefore expressed with dispro-
portionate intensity in nighttime rather than daytime tem-
peratures [Gershunov et al., 2009, hereinafter GCI09]. If this
trend continues, it will adversely affect Californias biota
acclimatized to the regions semi-arid Mediterranean climate
with its warm dry summer days and cool nights. Such a trend
translates into severe impacts on health, ecosystems, agricul-
ture, water resources, energy demand and infrastructure, all
with economic consequences. (For example, in late July 2006,
extreme heat with high humidity impacted human mortality
[Ostro et al., 2009] as well as morbidity [Knowlton et al.,
2009; Gershunov et al., 2011] and thus the healthcare indus-
try.) It is imperative to develop a detailed understanding of the
observed trend in regional heat wave activity if we are to
effectively mitigate its escalating impacts.
[
3] Californias complex topography and proximity to the
coast create distinct climate and ecological sub-regions within
relatively close distances. In different locations, regional-scale
heat waves may be expressed differently, or may respond to
different mechanisms altogether. For example, offshore
Santa Ana winds can cause coastal heat waves via adiabatic
warming of air as it descends from high elevations to sea
level. On the other hand, sea breeze and associated marine
layer clouds can moderate coastal temperatures during inland
heat waves. Furthermore, the non-uniform distribution of
population and resources across the state make for non-
uniform impacts. Although heat waves are synoptic phe-
nomena with a regional footprint, these complexities make a
sub-regional focus necessary.
[
4] We will examine observations and downscaled model
simulations to describe heat wave activity over six sub-regions
of California spanning the last six decades as well as that
projected for the 21st century. Before considering projections,
however, we screen GCMs for their ability to simulate
regional heat waves for the correct synoptic reasons. Station-
ary thresholds based on historical climate are used to quantify
sub-regional heat wave magnitudes in a changing climate.
This is the traditional approach often used in climate research
[Meehl et al., 2000; Tebaldi et al., 2006; Mastrandrea et al.,
2009; Diffenbaugh and Ashfaq, 2010]. Also, because climate
change is a long-term trend or non-stationarity in the daily
temperature climatology, we examine regional heat wave
activity relative to the contemporaneous climate, i.e., in a non-
stationary framework, which is relevant to estimating con-
temporaneous impacts of weather extremes.
2. Data and Methods
2.1. Observations
[
5] Observed data are daily maximum and minimum tem-
peratures (Tmax and Tmin, respectively) interpolated onto a
regular 12 12 km grid with temperature lapsed to grid cell
center elevations [Maurer et al., 2002]. The station source data
are from the National Climatic Data Center (NCDC) first-
order and cooperative observer summary of the day dataset
[National Climatic Data Center, 2003]. Daily sea level pres-
sure (SLP) and precipitable water (PRWTR) data are from
NCEP/NCAR reanalysis.
2.2. Model Simulations and Downscaling
[
6] Four GCMs were chosen (CNRM, GFDL, CCSM
and PCM) from those participating in the Coupled Model
Intercomparison Project Phase 3 (CMIP3) and Intergovern-
mental Panel on Climate Change 4th Assessment Report
(AR4) for their ability to simulate seasonal features of
Californias Mediterranean-type climate and observed cli-
mate variability as well as for availability of requisite vari-
ables at daily temporal resolution (see Text S1 in the
auxiliary material for details).
1
The spatially coarse GCM
information from the historical and SRES A2 simulations
1
Scripps Institution of Oceanography, University of California, San
Diego, La Jolla, California, USA.
Corresponding author: A. Gershunov, Scripps Institution of
Oceanography, University of California, San Diego, 9500 Gilman Dr.,
La Jolla, CA 92093, USA. (sasha@ucsd.edu)
©2012. American Geophysical Union. All Rights Reserved.
0094-8276/12/2012GL052979
1
Auxiliary materials are available in the HTML. doi:10.1029/
2012GL052979.
GEOPHYSICAL RESEARCH LETTERS, VOL. 39, L18710, doi:10.1029/2012GL052979, 2012
L18710 1of7
was statistically downscaled onto the 12 12 km grid
according to the methodology of Maurer et al. [2010], which
is based on a constructed analogues approach (CA) of Maurer
and Hidalgo [2008] matching modeled and observed daily
large-scale anomalies with a quantile-mapping bias correction
on the large-scale data prior to the CA approach. We use this
statistical downscaling approach with awareness of the implicit
assumption that the observed relationships between large-scale
and fine-scale weather are stationary under climate change.
2.3. Heat Wave Definition
[
7] Heat waves were defined as in GCI09 and Guirguis
et al. [2011] as daily temperature excesses over a threshold.
We use daily Tmax and Tmin exceeding their respective 95th
percentile (p95) thresholds computed from the local daily
climatologies over the May through September warm season
spanning the years 19501999. The heat wave index (HWI)
so defined captures the frequency, intensity and duration of
heat waves. The HWI is calculated daily and locally for each
12 km grid box. We obtain regional and/or seasonal heat
wave indices by aggregating daily values over a region or
season, respectively. The observed daily HWI values aver-
aged over the entire state are shown in Figure 1a.
2.4. Model Validation
[
8] As with most extremes, the rare nature of regional heat
waves requires that GCMs be validated for their ability to
realistically simulate them for the correct dynamical/synoptic
reasons. Validation of the four GCMs followed the synoptic
analysis of GCI09 to diagnose SLP anomalies that produce
extensive heat waves in California (Figure S1). Only one of the
models (CNRM) performed realistically prominently exhi-
biting both a surface High over the Great Plains and a Pacific
Low that together produce southerly advection over California
State. Other models are deficient in this regard displaying an
overly extensive Pacific Low that extends over California and
an underdeveloped Great Plains High. Daily PRWTR data
(uniquely available from the CNRM model) was further
verified to exhibit a perfectly realistic differentiation between
Type I and II heat waves increased PRWTR over the region
during the more humid Type II heat waves (not shown). This
CNRM version 3 GCM (CNRM [Salas-Mélia et al., 2005])
boasts a good overall performance in simulating southwest-
ern regional climate features [e.g., Favre and Gershunov,
2009; Das et al., 2011; Cayan et al., 2010]. We further
focus on the downscaled past (19502010) and future (2011
2099) CNRM model runs, but also include results from the
other models in the auxiliary material for completeness. The
past period was selected as a concatenation of the historical
and projected model simulations that is consistent with the
observational period.
2.5. California Regionalization
[
9] Unique California climate regions (Figure 2a, also
referred to as sub-regions) were identified using rotated
principal components analysis (PCA) to isolate spatially cohe-
sive regions experiencing similar temporal variability. Numer-
ous studies have used PCA for regionalization purposes [e.g.,
Richman and Lamb, 1985; Ehrendorfer, 1987; Comrie and
Glenn, 1998; Guirguis and Avissar, 2008]. We applied PCA
to the local daytime HWI over the 19501999 base period to
identify the primary patterns of summertime heat wave vari-
ability. Six PCs representing 65% of the overall variance
were varimax rotated and regions were identified following
the maximum loading method [Comrie and Glenn, 1998].
The results were stable to choice of base period and orthog-
onal versus oblique rotation as well as to whether the daytime
or nighttime HWI was used. We tested the ability of the
models to represent Californias sub-regional climatology by
repeating the regionalization using downscaled data and the
results were very similar, especially for CNRM, to observed
(not shown). Accordingly, the grid coordinates of observed
climate sub-regions were used for the regional analyses.
3. Results
3.1. State-Wide Heat Wave Activity
[
10] Figure 1 (Figure S2) shows the daily HWI averaged
over California in observations and the CNRM model (all
models). Strong interannual and decadal variability in heat
wave activity is evident in all data. In earlier decades, Type I
events dominated in frequency and intensity, but Type II
events have become more dominant in the last couple of
decades. Observations and model simulations are compara-
ble in this regard, although models tend to show a higher
recent propensity towards both types of events. Although
CNRM simulates the most realistic heat waves in terms of
synoptic forcing (Text S1), its HWI trend is overestimated
compared to observations and other models (Table 1). We
will not discuss the absolute magnitudes of these trends,
but rather concentrate on their qualitative manifestations,
noting that all models show significant increases in heat
wave activity of both types, although with a consistently
disproportionate increase in Type II heat waves in future
projections (Table 1).
Figure 1. Spatially-averaged state-wide heat wave index (HWI) derived independently from Tmax and Tmin (see text) for
(a) observations and (b) the CNRM model. An individual heat wave is defined as a group of consecutive days in which either
Tmin or Tmax exceeded the p95 threshold. A Type I heat wave (red) is one when the HWI sum for Tmax exceeded that for
Tmin. For Type II heat waves the Tmin HWI sum was greater (blue).
GERSHUNOV AND GUIRGUIS: CALIFORNIA HEAT WAVES PRESENT AND FUTURE L18710L18710
2of7
3.2. Regional Heat Wave Activity
[
11] The peak of the heat wave season typically occurs at
the seasonal temperature peak in late July for all inland
regions (Figures 2b and 2c). Coastal regions present an
exception to this rule, however, particularly for Type I heat
waves, with a bi-modal distribution and the primary peak in
September (also seen in Figures 3b and 3c). When inland
temperatures are at their seasonal maximum, the sea breeze
and associated marine layer clouds tend to cool the coastal
regions. The 95th percentile thresholds are therefore cooler
in coastal regions and are more readily exceeded early and
especially late in the season. In September, a regional kata-
batic offshore wind, known in Southern California as the
Santa Ana [Raphael, 2003; Hughes and Hall, 2010], causes
dry coastally trapped heat waves as high desert air descends
from the Great Basin (1000 m above sea level) and over
the coastal ranges, warming adiabatically and drying on its
way down to the coast. The historical GCM simulations
reproduce these late-season, predominantly Type I, coastal
heat waves, albeit with underestimated intensity (Figures 3e
and 3f). The observed coastal heat wave season is therefore
somewhat broader with more frequent and intense late-
season predominantly Type I events (Figures 3b and 3c; also
true for the other models, not shown).
[
12] Observed nighttime heat waves are seen to be signif-
icantly increasing in all regions (Table 1 and Figure 2e). In
coastal regions, midsummer nighttime heat waves have
shown a marked increase beginning in the 1980s (Figures 2e,
3b, and 3c). This coastal trend in midsummer nighttime heat
waves is due to an increase in spatial extent of inland-
centered heat events causing them to encroach upon coastal
areas (Figure 4), a development perhaps facilitated by greater
atmospheric stability at night and associated land-breeze
circulation. Modeled Type II heat waves are increasing more
than Type I for most regions, except along the coast where
both types tend to be comparably on the rise, broadly con-
sistent with observations (Table 1).
[
13] Although observed long-term trends in Type I heat
wave activity are positive for all regions, except the heavily
irrigated Central Valley, only the Coastal North shows a
significant trend in daytime heat waves (Table 1). Modeled
trends are also shown in Table 1 for the simulated past and
future climate periods. There is a significant modeled trend
Table 1. Table of Daytime and Nighttime Seasonal HWI Trends for Observations (Obs) and Downscaled GCMs Given in
C per Decade
a
Obs CNRM GFDL CCSM PCM
Daytime Nighttime Daytime Nighttime Daytime Nighttime Daytime Nighttime Daytime Nighttime
CA 0.4 1.6 3.7 (33.0) 3.9 (43.2) 2.5 (19.6) 2.3 (21.2) 1.1 (17.9) 1.2 (21.6) 1.5 (9.2) 1.3 (10.0)
Central Valley 0.9 1.6 3.8 (32.2) 4.3 (43.6) 2.4 (16.8) 2.4 (19.0) 0.9 (16.6) 1.1 (19.4) 1.6 (9.7) 1.5 (10.3)
Southern Deserts 0.6 2.5 3.1 (31.8) 3.4 (36.4) 2.1 (24.3) 2.6 (23.1) 1.8 (22.4) 1.2 (21.9) 1.2 (9.3) 1.4 (9.4)
Coastal North 2.0 1.7 4.5 (54.4) 3.6 (58.2) 2.1 (19.0) 1.5 (26.1) 1.5 (16.9) 1.5 (28.1) 2.1 (9.1) 1.3 (11.7)
Coastal South 1.1 2.3 3.7 (53.4) 3.6 (55.3) 2.1 (23.3) 1.7 (20.9) 1.6 (20.2) 1.3 (24.0) 2.1 (8.9) 2.0 (10.3)
Mojave 0.9 1.4 3.4 (25.5) 4.0 (38.0) 2.5 (23.0) 2.7 (21.7) 1.5 (21.1) 1.1 (22.1) 1.2 (9.4) 1.1 (10.1)
Northern Forests 0.3 1.2 3.7 (26.4) 3.9 (41.2) 3.1 (16.6) 2.3 (20.1) 0.5 (13.9) 1.1 (20.2) 1.2 (8.5) 0.8 (9.4)
a
Bold font indicates trends are significant at the 95% level. Trends for the historical period (19502010) and future (20112099) are provided for the
GCMs, with the future trends shown in parentheses.
Figure 2. (a) California sub-regions derived from daily HWI variability, (b, c) seasonality of observed daytime and night-
time heat waves for each California sub-region and (d, e) time series of the seasonal daytime and nighttime HWI shown
smoothed with a 5-point running mean. The colors in Figures 2b2e correspond to the sub-regions in Figure 2a.
GERSHUNOV AND GUIRGUIS: CALIFORNIA HEAT WAVES PRESENT AND FUTURE L18710L18710
3of7
for the past period in both daytime and nighttime heat waves.
Excepting the coastal regions, past nighttime trends tend to
be greater for most models. Future climate shows intensified
increasing trends, with trends in Type II heat waves exceed-
ing those of Type I for the vast majority of regions and
models (Table 1). Moreover, most models project both the
North and South Coasts to experience the largest increase in
Type II heat wave magnitude.
[
14] Because exceedances over local temperature thresh-
olds are accumulated over space (region) and time (season) a
general warming resulting in stronger and more persistent
exceedances over once-extreme thresholds leads to enor-
mous trends as these thresholds are rendered less and less
extreme and are exceeded on a more and more regular basis
in a warming climate. In other words, the inflated HWI trends
in future climate are due to stationary thresholds applied in a
non-stationary climate.
3.3. Non-stationary Variance Structure of Daily
Temperature
[
15] The magnitude of trends in heat wave activity,
although everywhere positive, varies by region (Table 1).
These regional variations of trends may be due to varying
magnitudes of heat waves relative to the regional back-
ground warming. To consider this possibility, we define rare
extremes relative to the typical values making up most of the
Figure 3. As in Figure 1 but for selected California sub-regions.
Figure 4. Longitudinal averages of the July nighttime HWI for three neighboring regions showing how the spatial extent
of Type II heat waves is increasing and their influence is spreading to coastal regions in July, the peak of the inland heat
wave season.
GERSHUNOV AND GUIRGUIS: CALIFORNIA HEAT WAVES PRESENT AND FUTURE L18710L18710
4of7
contemporary probability density function (PDF), i.e., tem-
peratures to which biota can potentially acclimate with
time. Thus, we next consider extremes relative to a changing
climatology.
[
16] To evaluate the intensity of extremes in this non-
stationary framework, we examine the ratio of p95 to the
median (p50) of daily Tmax and Tmin PDF in each summer
season. Maps and time series of p95-p50 (Figure 5) reflect
the magnitude of sub-regional heat waves relative to median
seasonal temperatures and therefore the severity of heat
waves relative to a non-stationary climate. Let us call this
index the relative heat index or RHI.
[
17] Although Type II heat wave activity increased and
was simulated and projected to increase in all sub-regions
(Table 1 and Figure 3e), observations show decreasing heat
wave magnitude relative to seasonal median warming (i.e.,
decreasing RHI) in the inland regions (Figure 5c). The
Southern and Mojave Deserts are the only regions where
RHI, for either type of heat waves, is projected by the CNRM
to significantly decrease relative to continuing median warm-
ing (Figures 5 and S4). In these hottest of the six regions, the
change in the magnitude of heat waves seems limited by that
of seasonal warming. Observations and projections, as well as
from the other models, seem to broadly agree on this point
(Figures 5 as well as Figures S3 and S4).
[
18] At the same time, we observe increasing heat wave
magnitude relative to seasonal median warming (i.e., increas-
ing RHI) only along the North Coast. This is the only region
where CNRM simulates significant change over the obser-
vational period an increase consistent with the observed.
The model furthermore projects increasing RHI exclusively
for both coastal regions (see Figures 5a5c for Type II and
Figure S4 for Type I). Other models tend to agree (Figure S3).
Summertime coastal seasonal warming is expected to be
modulated by cool coastal waters and is projected to be much
weaker than inland warming [Cayan et al., 2008, 2009].
Increasing frequency, intensity, duration and especially spatial
extent of large inland Type II heat waves (GCI09), however,
result in humid nighttime-accentuated heat encroaching
more and more into coastal regions, particularly at night when
the sea breeze weakens. This effect is already visible in the
observations over the northern part of the state (Figure 4). This
intensification of coastal heat waves relative to the slower
average warming explains the stronger projected coastal HWI
Figure 5. The relative heat index (RHI) index calculated at each grid cell from CNRM (a) maximum and (b) minimum
temperatures and shown as the difference between the mean RHI for a given climate period (20212050 or 20702099)
and (c) the mean RHI over the historical period (19502010) and the RHI index calculated from regionally aggregated
minimum temperatures according to observations (black) and CNRM simulations (red). Solid lines show linear trends
significant at the 95th confidence level calculated separately for the past (19502010) and future (20112100) periods.
GERSHUNOV AND GUIRGUIS: CALIFORNIA HEAT WAVES PRESENT AND FUTURE L18710L18710
5of7
trends relative to other regions (Table 1). The models pro-
jecting strong intensification of heat wave activity along the
coast relative to inland regions broadly agree on this point.
4. Discussion and Conclusions
[19] We examined observed and modeled heat wave
activity over California and its sub-regions. Four downscaled
GCM simulations were validated against observations for
their ability to reproduce heat waves in terms of realistic
synoptic forcings. The most realistic model was found to also
best reproduce the most important feature of the observed
regional trend the stronger relative trend in Type II (humid,
Tmin-accentuated) versus Type I (dry, Tmax-accentuated)
regional heat waves. This differential trend was most coher-
ent for the inland regions in observations and in most models.
Projected heat wave activity was examined in the context of
observed trends using stationary thresholds as well as in a
non-stationary context.
[
20] The analysis of heat waves involving stationary
thresholds, the 95th percentile of the local historical sum-
mertime climatology of daily temperatures, is relevant to
examining how current acclimatization and existing infra-
structure might perform in future heat waves. This conven-
tional approach suggests that, according to the GCMs, the
trend towards stronger and more frequent Type II relative to
Type I heat waves is a signal of anthropogenic climate
change that is expected to continue and accelerate through
the 21st century. However, much of this change in extremes
may be produced simply by the trend in the mean summer-
time temperature also observed to have been strongest in
Tmin rather than Tmax over this region [Gershunov and
Cayan, 2008] as well as over many other regions around
the globe [e.g., Easterling et al., 1997].
[
21] The past and future regional trends are positive as can
be expected, because, following convention, we have first
defined heat waves relative to stationary local thresholds in a
warming non-stationary climate. In other words, exceedance
of stationary historical thresholds is bound to increase over
time due simply to a trend in the mean of the daily temper-
ature distribution, so that a long-term mean warming will
eventually make historical heat waves commonplace. This
does not mean that heat wave activity per se is expected to
change relative to the changing mean climate. And it does not
explain why coastal heat waves are projected to change rel-
ative to stationary thresholds more than their inland coun-
terparts while coastal mean seasonal warming is projected to
be less than that inland [Cayan et al., 2008, 2009].
[
22] The non-stationary analysis provides insight into how
heat waves may change relative to seasonal warming, or how
the shape of the daily summertime Tmax and Tmin PDFs and
especially their tails may change relative to the warming mean
state. Non-stationary analysis suggests that heat waves in hot
desert regions will likely become less intense relative to strong
inland warming, especially at night, while coastal heat waves
will become more intense relative to the milder warming
modulated by the cool oceans proximity. This projected
coastal trend is enhanced for Type II heat waves in July and it
is already observed in coastal Northern California. It largely
represents coastal penetration of expanding Type II midsum-
mer inland heat waves.
[
23] This trend is particularly distressing for coastal regions
where population density is the highest, while physiological
acclimatization and air conditioning penetration are the lowest
for the state [Sailor and Pavlova, 2003]. The non-stationary
view is relevant for estimating future impacts on health,
ecosystems and agriculture as it implicitly incorporates the
effects of adaptation (e.g., via physiological or technologi-
cal acclimatization) to the evolving climate state. The strong
and coherent impact on public health, as measured by emer-
gency department visits, can already be detected most clearly
along the north-central coast when considering the unprece-
dented heat wave of July 2006 [Knowlton et al., 2009;
Gershunov et al., 2011]. Given results presented here, that
unprecedented Type II event can be considered a harbinger of
sub-regional impacts to come. In the future, little acclimati-
zation to warmer temperatures can be expected due to the
relatively weak warming projected along the highly popu-
lated California coast compared to inland areas [Cayan et al.,
2008, 2009]. Therefore, a disproportionate coastal increase
in heat wave activity is bound to be impactful. Coastal areas
will be the subject of more in-depth future research. In
particular, the interaction of nighttime land-breeze and late-
summer Santa Ana winds on the one hand with sea-breeze-
related coastal marine layer cloud influences on the other,
will need to be examined in detail to develop a refined and
rigorous physical understanding of coastal heat wave trends.
[
24] Acknowledgments. This work was supported by DOI via the
Southwest Climate Science Center, by NOAA via the RISA program
through the California and Nevada Applications Center, and by the
National Institute of Environmental Health Sciences grant RC1ES019073
(Projected heat wave magnitudes and public health impacts ; PI Margolis,
H.G.). Kristen Guirguis was supported via a UCAR PACE fellowship. We
thank Mary Tyree for data retrieval, downscaling and handling. The order
of authorship was determined by a coin toss.
[25] The Editor thanks the two anonymous reviewers for assisting in the
evaluation of this paper.
References
Cayan, D. R., E. P. Maurer, M. D. Dettinger, M. Tyree, and K. Hayhoe
(2008), Climate change scenarios for the California region, Clim.
Change, 87,2142, doi:10.1007/s10584-007-9377-6.
Cayan, D., M. Tyree, M. Dettinger, H. Hidalgo, T. Das, E. Maurer,
P. Bromirski, N. Graham, and R. Flick (2009), Climate change scenarios
and sea level rise estimates for the California 2009 Climate Change Sce-
narios Assessment, Publ. CEC-500-2009-014-F, 64 pp., Clim. Change
Cent., La Jolla, Calif.
Cayan, D. R., T. Das, D. W. Pierce, T. P. Barnett, M. Tyree, and A. Gershunov
(2010), Future dryness in the southwest US and the hydrology of the early
21st century drought, Proc. Natl. Acad. Sci. U. S. A., 107,21,27121,276,
doi:10.1073/pnas.0912391107.
Comrie, A. C., and E. C. Glenn (1998), Principal components based region-
alization of precipitation regimes across the southwest United States and
northern Mexico, with an application to monsoon precipitation variabil-
ity, Clim. Res., 10, 201215, doi:10.3354/cr010201.
Das, T., M. D. Dettinger, D. R. Cayan, H. G. Hidalgo (2011), Potential
increase in floods in Californias Sierra Nevada under future climate
projections. Climatic Change, 109, S71S94, doi:10.1007/s10584-011-
0298-z.
Diffenbaugh, N. S., and M. Ashfaq (2010), Intensification of hot extremes
in the United States, Geophys. Res. Lett., 37, L15701, doi:10.1029/
2010GL043888.
Easterling, D. R., et al. (1997), Maximum and minimum temperature trends
for the globe, Science, 277, 364367, doi:10.1126/science.277.5324.364.
Ehrendorfer, M. (1987), A regionalization of Austrias precipitation cli-
mate using principal component analysis, Int. J. Climatol. , 7,7189,
doi:10.1002/joc.3370070107.
Favre, A., and A. Gershunov (2009), North Pacific cyclonic and anticyclonic
transients in a global warming context: possible consequences for western
North American daily precipitation and temperature extremes, Clim. Dyn.,
32, 969987, doi:10.1007/s00382-008-0417-3.
Gershunov, A., and D. Cayan (2008), Recent increase in California heat
waves: July 2006 and the last six decades, PIER Project report, Calif.
Energy Comm., Sacramento. [Available at http://www.energy.ca.gov/
2008publications/CEC-500-2008-088/CEC-500-2008-088.PDF.]
GERSHUNOV AND GUIRGUIS: CALIFORNIA HEAT WAVES PRESENT AND FUTURE L18710L18710
6of7
Gershunov, A., D. Cayan, and S. Iacobellis (2009), The great 2006 heat
wave over California and Nevada: Signal of an increasing trend, J. Clim.,
22, 61816203, doi:10.1175/2009JCLI2465.1.
Gershunov, A., Z. Johnston, H. G. Margolis, and K. Guirguis (2011), The
California heat wave 2006 with impacts on statewide medical emer-
gency, Geogr. Res. Forum, 31,5369.
Guirguis, K. J., and R. Avissar (2008), A precipitation climatology and
dataset intercomparison for the western United States, J. Hydrometeorol.,
9, 825841, doi:10.1175/2008JHM832.1.
Guirguis, K., A. Gershunov, R. Schwartz, and S. Bennett (2011), Recent
warm and cold daily winter temperature extremes in the Northern Hemi-
sphere, Geophys. Res. Lett., 38, L17701, doi:10.1029/2011GL048762.
Hughes, M., and A. Hall (2010), Local and synoptic mechanisms causing
Southern Californias Santa Ana winds. Climate Dynamics, 34, 847857.
Knowlton, K., M. Rotkin-Ellman, G. King, H. G. Margolis, D. Smith,
G. Solomon, R. Trent, and P. English (2009), The 2006 California heat
wave: Impacts on hospitalizations and emergency department visits,
Environ. Health Perspect., 117,6167.
Mastrandrea, M. D., C. Tebaldi, C. P. Snyder, and S. H. Schneider (2009),
Current and future impacts of extreme events in California, PIER Tech.
Rep. CEC-500-2009-026-D, Calif. Energy Comm., Sacramento.
Maurer, E. P., and H. G. Hidalgo (2008), Utility of daily vs. monthly large-
scale climate data: An intercomparison of two statistical downscaling
methods, Hydrol. Earth Syst. Sci., 12, 551563, doi:10.5194/hess-12-
551-2008.
Maurer, E. P., A. W. Wood, J. C. Adam, D. P. Lettenmaier, and B. Nijssen
(2002), A long-term hydrologically-based data set of land surface fluxes
and states for the conterminous United States, J. Clim., 15, 32373251,
doi:10.1175/1520-0442(2002)015<3237:ALTHBD>2.0.CO;2.
Maurer, E. P., H. G. Hidalgo, T. Das, M. D. Dettinger, and D. R. Cayan
(2010), The utility of daily large-scale climate data in the assessment of
climate change impacts on daily streamflow in California, Hydrol. Earth
Syst. Sci., 14, 11251138, doi:10.5194/hess-14-1125-2010.
Meehl, G. A., et al. (2000), An introduction to trends in extreme weather
and climate events: Observations, socioeconomic impacts, terrestrial eco-
logical impacts, and model projections, Bull. Am. Meteorol. Soc., 81,
413416, doi:10.1175/1520-0477(2000)081<0413:AITTIE>2.3.CO;2.
National Climatic Date Center (2003), Data documentation for data set
3200 (DSI-3200): Surface land daily cooperative summary of the day,
report, Asheville, N. C. [Available at http://www.ncdc.noaa.gov/pub/
data/documentlibrary/tddoc/td3200.pdf.]
Ostro, B. D., L. A. Roth, R. S. Green, and R. Basu (2009), Estimating the
mortality effect of the July 2006 California heat wave. Environ. Res.,
109
, 614619.
Raphael M. N. (2003), The Santa Ana winds of California. Earth Interactions,
7,113.
Richman, M. B., and P. J. Lamb (1985), Climatic pattern analysis of three-
and seven-day summer rainfall in the central United States: Some meth-
odological considerations and a regionalization, J. Clim. Appl. Meteorol.,
24, 13251343, doi:10.1175/1520-0450(1985)024<1325:CPAOTA>2.0.
CO;2.
Sailor, D., and A. Pavlova (2003), Air conditioning market saturation and
long-term response of residential cooling energy demand to climate
change, Energy, 28, 941951, doi:10.1016/S0360-5442(03)00033-1.
Salas-Mélia, D., F. Chauvin, M. Déqué, H. Douville, J. F. Guérémy,
P. Marquet, S. Planton, J. F. Royer, and S. Tyteca (2005), Description
and validation of the CNRM-CM3 global coupled model, CNRM Work.
Note 103, Cent. Natl. de Rech. Meteorol., Toulouse, France.
Tebaldi, C., K. Hayhoe, and J. M. Arblaster (2006), Going to the extremes:
an intercomparison of model-simulated historical and future changes in
extreme events, Clim. Change, 79, 185211, doi:10.1007/s10584-006-
9051-4.
GERSHUNOV AND GUIRGUIS: CALIFORNIA HEAT WAVES PRESENT AND FUTURE L18710L18710
7of7
... Most studies that address this challenge have used some variation of statistical downscaling (e.g. Miller et al., 2008;Gershunov and Guirguis, 2012;Xu et al., 2012). Building energy modeling of future extreme heat with dynamical downscaling has also been performed in the literature. ...
Article
While projected urban air temperatures under climate change and urbanization have received attention, projections of pedestrian thermal stress are scarce and usually produced by statistical downscaling. In this study, we present and evaluate a dynamical downscaling methodology to assess urban outdoor thermal exposure. We dynamically downscale Earth system model output with a mesoscale weather, urban canopy and building energy model (WRF-BEP-BEM), then use the downscaled output as boundary conditions for a microscale model (TUF-Pedestrian) to determine metre-scale mean radiant temperature. Using this methodology, we assess outdoor heat exposure and air conditioning loading during typical heatwaves at the start and end of century in Phoenix and Toronto. Results reveal that air conditioning loading would double in Phoenix by the end of century under an RCP 8.5 climate and urban development scenario; and it may double in Toronto even without urban development. This is compounded by up to 2 additional hours/day of extreme heat stress in Phoenix (5.5 additional hours/day of strong heat stress in Toronto). Using our dynamical downscaling methodology, we find that projected climate change-induced increases in outdoor heat stress largely result from higher air temperatures (Phoenix: 69-87%; Toronto: 60-64%), with smaller contributions from increased mean radiant temperature and humidity.
... In addition to the projected increases in the magnitude of mean, as well as extreme, temperatures, the changing persistence of temperature extremes (i.e., heat waves) is of grave concern. Perkins [3] describes heat waves as longer-than-normal periods of high temperatures that, in addition to wider contrasts between short-term and seasonal minimum and maximum temperatures, can significantly impact the natural and built environments (e.g., structural integrity of bridges and buildings and damage to pavements and railways) as well as have serious negative implications on human health and mortality [4]. Multiple studies have assessed past changes in the behavior of heat waves over the last several decades in the US [5] and globally [3,6]. ...
Article
Full-text available
In order to better understand the extent to which global climate variability is linked to the frequency and intensity of heat waves and overall changes in temperature throughout the United States (US), correlations between long-term monthly mean, minimum, and maximum temperatures throughout the contiguous US on the one hand and low-frequency variability of multiple climate indices (CIs) on the other hand are analyzed for the period from 1948 to 2018. The Pearson’s correlation coefficient is used to assess correlation strength, while leave-one-out cross-validation and a bootstrapping technique (p-value) are used to address potential serial and spurious correlations and assess the significance of each correlation. Three parameters defined the sliding windows over which surface temperature and CI values were averaged: window size, lag time between the temperature and CI windows, and the beginning month of the temperature window. A 60-month sliding window size and 0 lag time resulted in the highest correlations overall; beginning months were optimized on an individual site basis. High (r ≥ 0.60) and significant (p-value ≤ 0.05) correlations were identified. The Western Hemisphere Warm Pool (WHWP) and El Niño/Southern Oscillation (ENSO) exhibited the strongest links to temperatures in the western US, tropical Atlantic sea surface temperatures to temperatures in the central US, the WHWP to temperatures throughout much of the eastern US, and atmospheric patterns over the northern Atlantic to temperatures in the Northeast and Southeast. The final results were compared to results from previous studies focused on precipitation and coastal sea levels. Regional consistency was found regarding links between the northern Atlantic and overall weather and coastal sea levels in the Northeast and Southeast as well as on weather in the upper Midwest. Though the MJO and WHWP revealed dominant links with precipitation and temperature, respectively, throughout the West, ENSO revealed consistent links to sea levels and surface temperatures along the West Coast. These results help to focus future research on specific mechanisms of large-scale climate variability linked to US regional climate variability and prediction potential.
... Hotter temperatures from a warming climate coupled with lower moisture from changes in precipitation regimes also produce drier conditions, intensifying the risk for larger wild res 9,10 . Wild res and heat waves have increased in length, intensity, and size under climate change, particularly in the US West Coast [11][12][13][14][15] , with record-breaking events occurring frequently in recent years 2 . ...
Preprint
Full-text available
Extreme heat and wildfire smoke events are increasingly co-occurring in the context of climate change, especially in California. Extreme heat and wildfire smoke may have synergistic effects on population health that vary over space. We leveraged high-resolution satellite and monitoring data to quantify spatially varying compound exposures to extreme heat and wildfire smoke in California (2006–2019) at ZIP code level. We found synergistic effects between extreme heat and wildfire smoke on cardiorespiratory hospitalizations at the state level. We also found spatial heterogeneity in such synergistic effects across ZIP codes. Communities with lower education attainment, lower health insurance coverage, lower income, lower proportion of automobile ownership, lower tree canopy coverage, higher population density, and higher proportions of racial/ethnic minorities are more vulnerable to the synergistic effects. This study highlights the need to incorporate compound hazards and environmental justice considerations into evidence-based policy development to protect populations from increasingly prevalent compound hazards.
... In addition to the projected increases in the magnitude of mean, as well as extreme, temperatures, the changing persistence of temperature extremes (i.e., heat waves) is of grave concern. Perkins [3] describes heat waves as longer-than-normal periods of high temperatures that, in addition to wider contrasts between short-term and seasonal minimum and maximum temperatures, can significantly impact the natural and built environments (e.g., structural integrity of bridges and buildings and damage to pavements and railways) as well as have serious negative implications on human health and mortality [4]. Multiple studies have assessed past changes in the behavior of heat waves over the last several decades in the US [5] and globally [3,6]. ...
Preprint
Full-text available
In order to better understand the extent to which global climate variability is linked to the frequency and intensity of heat waves and overall changes in temperature throughout the United States (US), correlations between long-term monthly mean, minimum, and maximum temperatures measured at sites throughout the contiguous US and low-frequency variability of multiple climate indices (CIs) are analyzed over the period 1948 to 2018 using correlation analysis. The Pearson’s correlation coefficient is used to assess correlation strength, while Leave-One-Out Cross-Validation and a bootstrapping technique (p-value) are used to address potential serial and spurious correlation and assess the significance of each correlation. Three parameters defined the sliding windows over which surface temperature and CI values were averaged: window size, lag time between the temperature and CI windows, and the beginning month of the temperature window. A 60-month sliding window size and 0 lag time resulted in the strongest correlations overall; beginning months were optimized on an individual site basis. Strong (r >= 0.60) and significant (p-value <= 0.05) correlations were identified. The Western Hemisphere Warm Pool (WHWP) and El Niño/Southern Oscillation (ENSO) exhibited the strongest links to temperatures in the western US, tropical Atlantic sea surface temperatures to temperatures in the central US, the WHWP to temperatures throughout much of the eastern US, and atmospheric patterns over the northern Atlantic to temperatures in the Northeast and Southeast. The final results were compared to results from previous studies focused on precipitation and coastal sea levels. Regional consistency was found regarding links between the northern Atlantic and overall weather and coastal sea levels in the Northeast and Southeast as well as on weather in the upper Midwest. Though the MJO and WHWP revealed dominant links with precipitation and temperature, respectively, throughout the West, ENSO revealed consistent links to sea levels and surface temperatures along the West Coast. These results help focus future research regarding specific mechanisms of climate variability that appear the exhibit strong links to US regional weather and sea level variability and prediction.
... Perhaps surprising to many familiar with the mild climate coasts and hot, dry inlands of California, it is the temperature buffering effects near the coast that are most threatened. Coastal southern California specifically is anticipated to experience the most substantial warming across all of California with greater humidity and higher nighttime temperatures coupled with more extreme heat events (Gershunov and Guirguis 2012). Warmer temperatures can affect already strained water resources. ...
... Therefore, urban parks, especially those with significant tree canopy cover, can provide temperature relief for people who can access these spaces during high temperature events. Recent downscaling of global climate models to regional climates in California suggests that the frequency and duration of heat waves is likely to increase in coming decades, especially in coastal areas (Gershunov and Guirguis 2012). From a planning point of view, these findings suggests the need for increasing shade planting in urban parks and ensuring equitable access to parks and other vegetated public spaces for real-time in situ cooling. ...
Article
Full-text available
Mobile bicycle transects were conducted in Golden Gate Park and surrounding residential neighborhoods of San Francisco, CA, during the period July–October 2012. Measurements included air temperature at six heights from 0.15 to 2.5 m, as well as relative humidity, solar radiation, and surface radiometric temperature. Park temperatures were lower than surrounding neighborhoods at all measurement levels, and mean Park Cool Island (PCI) values ranged from 0.9 °C for grass playing fields to 1.9 °C for tree groves during typical cool (~ 16 °C) summer conditions. These increased to 2.6 °C and 3.7 °C, respectively, during periodic warm events (~ 30 °C). Grass fields produced strong temperature lapse profiles, similar to asphalt surfaces, while dense trees produced a near-isothermal profile. Shade enhanced cooling by 1–8 °C throughout the 2.5-m layer, peaking near the surface, and frequently producing inversions below 0.5 m. Complexity in the lower surface layer temperature profile suggests measurements of air temperature at one height may be insufficient to characterize the PCI. The park also produced a consistent humidity enhancement, with differences in vapor pressure between urban and park surfaces of up to 10 hPa. The humidity difference was positively correlated with the PCI and also doubled between cool and warm periods. This, along with assessment of the role of shade, and the daytime peak, suggests evaporative cooling and shade cooling are the main drivers of the PCI in Golden Gate Park. The advection of cool air from the park into leeward neighborhoods was limited below 2.5 m, and was only observable within the first few hundred meters. Implications for urban planning suggest urban green spaces that provide abundant vegetation, including dense tree canopies, and are widely accessible to inhabitants can be most effective for in situ thermal comfort, particularly temperature relief during heat events.
... Inland provinces experienced fewer heatwaves of longer duration and greater intensity with DTR peaking at 17.9 °C in the North West. The national spatial variability of heatwave duration, frequency and intensity could be attributed to the cooling effect that the sea breeze and associated marine layer clouds have on coastal region [20]. Overall, the highest frequency of heatwaves occurred during the austral summer accounting for a total of 150 heatwave events out of 270 events that occurred during 2014 to 2019 (5-year period). ...
Article
Full-text available
Heatwaves can have severe impacts on human health extending from illness to mortality. These health effects are related to not only the physical phenomenon of heat itself but other characteristics such as frequency, intensity, and duration of heatwaves. Therefore, understanding heatwave characteristics is a crucial step in the development of heat-health warning systems (HHWS) that could prevent or reduce negative heat-related health outcomes. However, there are no South African studies that have quantified heatwaves with a threshold that incorporated a temperature metric based on a health outcome. To fill this gap, this study aimed to assess the spatial and temporal distribution and frequency of past (2014 – 2019) and future (period 2020 – 2039) heatwaves across South Africa. Heatwaves were defined using a threshold for diurnal temperature range (DTR) that was found to have measurable impacts on mortality. In the current climate, inland provinces experienced fewer heatwaves of longer duration and greater intensity compared to coastal provinces that experienced heatwaves of lower intensity. The highest frequency of heatwaves occurred during the austral summer accounting for a total of 150 events out of 270 from 2014 to 2019. The heatwave definition applied in this study also identified severe heatwaves across the country during late 2015 to early 2016 which was during the strongest El Niño event ever recorded to date. Record-breaking global temperatures were reported during this period; the North West province in South Africa was the worst affected experiencing heatwaves ranging from 12 to 77 days. Future climate analysis showed increasing trends in heatwave events with the greatest increases (80%—87%) expected to occur during summer months. The number of heatwaves occurring in cooler seasons is expected to increase with more events projected from the winter months of July and August, onwards. The findings of this study show that the identification of provinces and towns that experience intense, long-lasting heatwaves is crucial to inform development and implementation of targeted heat-health adaptation strategies. These findings could also guide authorities to prioritise vulnerable population groups such as the elderly and children living in high-risk areas likely to be affected by heatwaves.
Preprint
Full-text available
Extreme weather and climate events can have substantial impacts on society and the environment. Compound extremes (two or more extreme events occurring simultaneously or successively) may exert even larger impacts than individual events. This study examines physical drivers behind variability in hydrometeorological (precipitation and temperature) compound extremes on subseasonal-to-seasonal (S2S) timescales (2 weeks – 6 months). Observational evidence presented here indicates significant modulation of western U.S. compound extreme frequency by the Madden-Julian oscillation (MJO), a unique type of organized tropical convection varying primarily on S2S timescales. For example, when the MJO is active over the western Pacific, a robust increase in wet-cold extreme frequency is found in Southern California. When the MJO is over the Maritime Continent, an overall increase in dry-hot extremes is observed across the western U.S.. The MJO influence on compound extremes is largely modulated by El Niño/Southern Oscillation (ENSO), which can be seen through different magnitudes or changes in sign of the canonical MJO-extreme relationship conditioned on ENSO state. Similarly, the MJO can interrupt the canonical ENSO-compound extreme relationship. Our results suggest a potential route to improve western U.S. S2S prediction of compound hydrometeorological extremes by considering the combined effect of both MJO and ENSO.
Article
Heatwaves in California manifest as both dry and humid events. While both forms have become more prevalent, recent studies have identified a shift toward more humid events. Understanding the complex interactions of each heatwave type with the urban heat island is crucial for impacts but remains understudied. Here, we address this gap by contrasting how dry versus humid heatwaves shape the intraurban heat of the greater Los Angeles area. We used a consecutive contrasting set of heatwaves from 2020 as a case study: a prolonged humid heatwave in August and an extremely dry heatwave in September. We used MERRA-2 reanalysis data to compare mesoscale dynamics, followed by high-resolution Weather Research and Forecasting modeling over urbanized Southern California. We employ moist thermodynamic variables to quantify heat stress and perform spatial clustering analysis to characterize the spatiotemporal intraurban variability. We find that, despite temperatures being 10° ± 3°C hotter in the September heatwave, the wet-bulb temperature, closely related to the risk of human heat stroke, was higher in August. While dry and humid heat display different spatial patterns, three distinct spatial clusters emerge based on nonheatwave local climates. Both types of heatwaves diminish the intraurban heat stress variability. Valley areas such as San Bernardino and Riverside experience the worst impacts, with up to 6° ± 0.5°C of additional heat stress during heatwave nights. Our results highlight the need to account for the disparity in small-scale heatwave patterns across urban neighborhoods in designing policies for equitable climate action. Significance Statement Heatwaves are the leading cause of morbidity and mortality among all environmental hazards. Moreover, their frequency and intensity are on the rise due to climate change. Southern California is no stranger to extreme heat, but persistently humid heatwaves still test the adaptability limits of its residents. We find that the set of two contrasting heatwaves that afflicted Los Angeles in the summer of 2020 forms the perfect testbed for characterizing the impacts of humid versus dry heatwaves on urban environment. Because climate model forecasts and long-term observational trends point to more humid heatwaves in the future for Southern California, our results underscore the importance of including moist heat in extreme heat warning frameworks.
Article
Full-text available
Projections of changes in climate extremes are critical to assessing the potential impacts of climate change on human and natural systems. Modeling advances now provide the opportunity of utilizing global general circulation models (GCMs) for projections of extreme temperature and precipitation indicators. We analyze historical and future simulations of ten such indicators as derived from an ensemble of 9 GCMs contributing to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC-AR4), under a range of emissions scenarios. Our focus is on the consensus from the GCM ensemble, in terms of direction and significance of the changes, at the global average and geographical scale. The climate extremes described by the ten indices range from heat-wave frequency to frost-day occurrence, from dry-spell length to heavy rainfall amounts. Historical trends generally agree with previous observational studies, providing a basic sense of reliability for the GCM simulations. Individual model projections for the 21st century across the three scenarios examined are in agreement in showing greater temperature extremes consistent with a warmer climate. For any specific temperature index, minor differences appear in the spatial distribution of the changes across models and across scenarios, while substantial differences appear in the relative magnitude of the trends under different emissions rates. Depictions of a wetter world and greater precipitation intensity emerge unequivocally in the global averages of most of the precipitation indices. However, consensus and significance are less strong when regional patterns are considered. This analysis provides a first overview of projected changes in climate extremes from the IPCC-AR4 model ensemble, and has significant implications with regard to climate projections for impact assessments.
Article
Full-text available
Projections of changes in climate extremes are critical to assessing the potential impacts of climate change on human and natural systems. Modeling advances now provide the opportunity of utilizing global general circulation models (GCMs) for projections of extreme temperature and precipitation indicators. We analyze historical and future simulations of ten such indicators as derived from an ensemble of 9 GCMs contributing to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC-AR4), under a range of emissions scenarios. Our focus is on the consensus from the GCM ensemble, in terms of direction and significance of the changes, at the global average and geographical scale. The climate extremes described by the ten indices range from heat-wave frequency to frost-day occurrence, from dry-spell length to heavy rainfall amounts. Historical trends generally agree with previous observational studies, providing a basic sense of reliability for the GCM simulations. Individual model projections for the 21st century across the three scenarios examined are in agreement in showing greater temperature extremes consistent with a warmer climate. For any specific temperature index, minor differences appear in the spatial distribution of the changes across models and across scenarios, while substantial differences appear in the relative magnitude of the trends under different emissions rates. Depictions of a wetter world and greater precipitation intensity emerge unequivocally in the global averages of most of the precipitation indices. However, consensus and significance are less strong when regional patterns are considered. This analysis provides a first overview of projected changes in climate extremes from the IPCC-AR4 model ensemble, and has significant implications with regard to climate projections for impact assessments.
Article
Full-text available
This paper presents the results of climatic pattern analyses of three- and seven-day summer (May-August) rainfall totals for the central US. A range of eigenvectorial methods is applied to 1949- 80 data for a regularly spaced network of 402 stations that extends from the Rocky to the Appalachian Mountains and from the Gulf Coast to the Canadian border. The major objectives are to quantitatively assess the sensitivity of eigenvectorial results to several parameters that have hitherto been the subject of considerable qualitative concern, and to identify the potential applications of those results. The entire domain variance fractions cumulatively explained by 1) the fist 10 correlation-based unrotated Principal Components (PCs) and 2) the 10 orthogonally rotated (VARIMAX criterion) PCs derived from them are identical for the same data. They vary between 35-47% depending on the data time scale and form, being higher for seven- than three-day totals and further enhanced when those totals are square-root (especially) and log 10 transformed. The (highly contrasting) sets of unrotated and VARIMAX PC spatial loading patterns are invariant with respect to data time scale and form. -from Authors
Article
Full-text available
This paper presents the results of a regionalization study of the precipitation climate of the western United States using principal component analysis. Past eigen-based regionalization studies have relied on rain gauge networks, which is restrictive because rain gauge coverage is sparse, especially over complex terrain that exists in the western United States. Here, the use of alternate data products is examined by conducting a comparative regionalization using nine precipitation datasets used in hydrometeorological research. Five unique precipitation climates are identified within the western United States, which have centers and boundaries that are physically reasonable and that highlight the relationship between the precipitation climatology and local topography. Using the congruence coefficient as the measure of similarity between principal component solutions, the method is found to be generally stable across datasets. The exception is the National Centers for Environmental Prediction-Department of Energy (NCEP-DOE) Reanalysis 2, which frequently demonstrates only borderline agreement with the other datasets. The loading pattern differences among datasets are shown to be primarily a result of data differences in the representation of (i) precipitation over the Rocky Mountains, (ii) the eastward wet-to-dry precipitation gradient that occurs during the cold season, (iii) the magnitude and spatial extent of the North American monsoon signal, and (iv) precipitation in the desert southwest during spring and summer. Sensitivity tests were conducted to determine whether the spatial resolution and temporal domain of the input data would dramatically affect the solution, and these results show the methodology to be stable to differences in spatial/temporal data features. The results suggest that alternate data products can be used in regionalization studies, which has applications for rain gauge installation and planning, climate research, and numerical modeling experiments.
Article
The unprecedented July 2006 California heat wave was associated with an accelerating trend in regional humid heat wave activity that is related to global warming. In this work, we identify and summarize effects of that meteorological extreme event on California morbidity as quantified by daily emergency department (ED) visits data aggregated at the county level. Besides local county-level daily correlation analysis, we use canonical correlation analysis (CCA) to summarize spacetime relationships between ED visits and temperature as well as humidity. We find intriguing spatial patterns of public health impacts of heat and humidity that potentially suggest modulation by various factors including demographics, adaptation and acclimatization that may be related to local climatology and social factors determining access to air conditioning. Health outcomes as measured by ED visits are found to be generally more sensitive to extreme heat and humidity in coastal counties and the north-central part of the Central Valley, which are known to be more affected by coastal influences, than in inland counties.
Article
Most of the great California-Nevada heat waves can be classified into primarily daytime or nighttime events depending on whether atmospheric conditions are dry or humid. A rash of nighttime-accentuated events in the last decade was punctuated by an unusually intense case in July 2006, which was the largest heat wave on record (1948-2006). Generally, there is a positive trend in heat wave activity over the entire region that is expressed most strongly and clearly in nighttime rather than daytime temperature extremes. This trend in nighttime heat wave activity has intensified markedly since the 1980s and especially since 2000. The two most recent nighttime heat waves were also strongly expressed in extreme daytime temperatures. Circulations associated with great regional heat waves advect hot air into the region. This air can be dry or moist, depending on whether a moisture source is available, causing heat waves to be expressed preferentially during day or night. A remote moisture source centered within a marine region west of Baja California has been increasing in prominence because of gradual sea surface warming and a related increase in atmospheric humidity. Adding to the very strong synoptic dynamics during the 2006 heat wave were a prolonged stream of moisture from this southwestern source and, despite the heightened humidity, an environment in which afternoon convection was suppressed, keeping cloudiness low and daytime temperatures high. The relative contributions of these factors and possible relations to global warming are discussed.
Article
A 33-yr, numerical dataset of the occurrence of Santa Ana winds for the period 1968–2000 has been created and validated. Daily Weather Maps were examined to identify the days when a surface high pressure system existed over the Great Basin simultaneously with a surface low pressure system offshore of southern California, and the prevailing wind over southern California was from the northeast quadrant. The dates of these occurrences, as well as the wind speed, temperature, and dewpoint temperature among other variables, were extracted and tabulated. The frequency of Santa Ana events derived from the weather maps was compared to events defined by wind direction only and there is agreement between the two. Preliminary results show that the Santa Ana event is limited to the period September–April and that the month of peak occurrence is December. The average frequency of events is 20 yr–1 and the average duration of an event is 1.5 days. Humidity levels are not uniform across Santa Ana events; the driest months are the months with the highest frequency of events. The frequency of Santa Ana events is usually lower than average during El Niño events. These preliminary results indicate that the dataset is useful for in depth study of the local phenomenon and its effect on the region within the context of the large-scale circulation.
Article
We determine precipitation regions for the United States-Mexico border region based on seasonality and variability of monthly precipitation at 309 stations for the period 1961 to 1990. Using a correlation matrix of input data to avoid the effect of elevation on precipitation, we apply principal components analysis with oblique rotation to regionalize this large, climatologically complex study area. We examine the applicability of the method, 2 techniques for defining region boundaries, the various defined regions themselves, and the effects of transforming input data and changing obliquity of component rotation. We obtain 9 consistent and largely contiguous regions from each of the analyses, including regions for the North American monsoon, the low deserts, the California Mediterranean regime, and for summer precipitation regimes adjoining the Gulf of Mexico. The derived regions and associated boundaries make physical sense in terms of the driving atmospheric processes, and they are robust to transformed input data and changes in rotation procedures. The central border regions are remarkably consistent across analyses, with small changes to peripheral regions. We also identify 4 monsoon sub-regions, and we illustrate the applicability of the regionalization via an analysis of relationships between monsoon precipitation variability and 500 mb pressure heights. Significantly different 500 mb circulation patterns are associated with wet and dry monsoon seasons in each of the sub-regions, and it appears that shifts in 500 mb circulation relative to the geographic position of each sub-region influence seasonal precipitation variability, directly or indirectly. There are important differences between some sub-regions, but in general wet monsoons are associated with northward meridional bulging of the subtropical anticyclone over the continental monsoon areas, while dry monsoons are associated with zonal stretching of the subtropical anticyclone over adjacent oceans with slightly higher pressure-heights. Overall, the study provides a clear regionalization of the precipitation climatology for the southwest United States and northern Mexico, and shows its utility for studies of climate variability.