A Statistical Analysis and Synoptic Climatology of Heat Waves over the Northeast United States Scott C. Runyon and Lance F. Bosart Department of Earth and Atmospheric Sciences, University at Albany, State University of New York

Overview/Motivation Heat waves are a major contributor to weather-related fatalitiesHeat waves are a major contributor to weather-related fatalities From 1986 to 2004 across the U.S.:From 1986 to 2004 across the U.S.: Source: Severe weather event # of fatalities Hurricanes*323 Tornadoes960 Lightning1081 Floods1575 Extreme heat 2596

Overview/Motivation (cont.) Improved forecasting may result from a better understanding of conditions antecedent to and during heat wavesImproved forecasting may result from a better understanding of conditions antecedent to and during heat waves These improved forecasts may become critical given the possibility of an increase in frequency & intensity of heat wavesThese improved forecasts may become critical given the possibility of an increase in frequency & intensity of heat waves

Overview/Motivation (cont.) Previous work has largely focused only on individual events (e.g., McQueen et al. 1956, 1957; Livezey and Tinker 1996) or on extended “heat wave–droughts” (e.g., Namias 1982, 1991)Previous work has largely focused only on individual events (e.g., McQueen et al. 1956, 1957; Livezey and Tinker 1996) or on extended “heat wave–droughts” (e.g., Namias 1982, 1991) Published synoptic climatologies have been limited in scope to “Midwest” or Great Plains (e.g., Namias 1955, 1983; Chang & Wallace 1984)Published synoptic climatologies have been limited in scope to “Midwest” or Great Plains (e.g., Namias 1955, 1983; Chang & Wallace 1984)

Goals Resolve annual and decadal trends in heat wave frequency nationwideResolve annual and decadal trends in heat wave frequency nationwide Investigate conditions that lead to onset of Northeast heat wavesInvestigate conditions that lead to onset of Northeast heat waves Understand synoptic evolution of and dynamic contributions to Northeast heat wavesUnderstand synoptic evolution of and dynamic contributions to Northeast heat waves

Overview MethodologyMethodology –Initial –Revised ResultsResults –Northeast vs. regional heat wave statistics –Northeast summer heat wave composites Antecedent conditionsAntecedent conditions Evolution of heat wavesEvolution of heat waves ConclusionsConclusions

Methodology Temperature data were gathered from the National Climatic Data Center’s (NCDC) high resolution surface datasetTemperature data were gathered from the National Climatic Data Center’s (NCDC) high resolution surface dataset Daily high temperatures were extracted for 54 surface stations over a 54-year period (1948–2001)Daily high temperatures were extracted for 54 surface stations over a 54-year period (1948–2001) Stations selected on basis of both dataset continuity and coverage within the NCDC’s Standard Regions for temperature and precipitationStations selected on basis of both dataset continuity and coverage within the NCDC’s Standard Regions for temperature and precipitation

Methodology (cont.) An anomalously hot day was initially defined as a day having a high temperature ≥ 2 standard deviations (σ) above the normal high temperature*An anomalously hot day was initially defined as a day having a high temperature ≥ 2 standard deviations (σ) above the normal high temperature* * 31-day running mean Result: Result: –Large variability in number of anomalous warm events between stations –Extreme ratios of anomalously hot days to anomalously cold days at many stations

Methodology (cont.) This initial method led to a discovery:This initial method led to a discovery: High temperatures at most stations are not normally distributed (“skewed”) Negatively Skewed Positively Skewed DEN ERI BOS LGA ALB CAR

Methodology (cont.) Temperature data is widely assumed to be normally distributedTemperature data is widely assumed to be normally distributed Los Angeles, CA and Denver, CO:Los Angeles, CA and Denver, CO: –Most positively and negatively skewed datasets, respectively –Using 2σ method 989 (49) anomalously hot days were found in Los Angeles (Denver) for all seasons

Most Positively Skewed Station: Los Angeles, CA Methodology (cont.)

Los Angeles, CA: Daily High Temperatures June 1 – August 31, 1948 – 2001 Mean Mean +2σ Mean −2σ Methodology (cont.)

Los Angeles, CA: Composite Mean Sea Level Pressure (hPa) – Ten Most Anomalous Warm Summer Days

Most Negatively Skewed Station: Denver, CO Methodology (cont.)

June 1 – August 31, 1948 – 2001 Mean Mean +2σ Mean −2σ Denver, CO: Daily High Temperatures

Denver, CO: Composite Mean Sea Level Pressure (hPa) – Ten Most Anomalous Cold Summer Days

Los Angeles:Los Angeles: –Large temperature anomalies more likely positive and associated with Santa Ana (offshore) events –Typically, marine layer moderates daily high temperatures Denver:Denver: –Large temperature anomalies more likely negative and associated with upslope events –Typically, mean westerly, downsloping flow influences daily high temperatures Non-normal temperature distributions are common and have distinct physical explanationsNon-normal temperature distributions are common and have distinct physical explanations Methodology (cont.)

Anomalous hot day: A day having a high temperature greater than the daily climatological 97.5 percentile thresholdAnomalous hot day: A day having a high temperature greater than the daily climatological 97.5 percentile threshold New method eased distribution problems [Los Angeles 669, Denver 736]New method eased distribution problems [Los Angeles 669, Denver 736] Heat wave: Three or more consecutive anomalously hot daysHeat wave: Three or more consecutive anomalously hot days Regional heat wave: When heat waves occurring at two or more stations within a region had at least one overlapping dayRegional heat wave: When heat waves occurring at two or more stations within a region had at least one overlapping day

Statistical Analysis

Statistical Analysis (cont.)

40 stations

Northeast Heat Wave Composites Created using NCEP/NCAR Reanalysis Dataset (0000 & 1200 UTC)Created using NCEP/NCAR Reanalysis Dataset (0000 & 1200 UTC)

Loop 1: Antecedent Conditions Composites are three-day averages centered on 7, 5, 3, & 1 day(s) prior to heat wave, 2 nd day into heat waveComposites are three-day averages centered on 7, 5, 3, & 1 day(s) prior to heat wave, 2 nd day into heat wave

Antecedent Conditions (7 Days Prior) n= hPa heights & anomalies (dam) 700 hPa heights & anomalies (dam)

Antecedent Conditions (5 Days Prior) n= hPa heights & anomalies (dam) 700 hPa heights & anomalies (dam)

Antecedent Conditions (3 Days Prior) n= hPa heights & anomalies (dam) 700 hPa heights & anomalies (dam)

Antecedent Conditions (1 Day Prior) n= hPa heights & anomalies (dam) 700 hPa heights & anomalies (dam)

Antecedent Conditions (Onset) n= hPa heights & anomalies (dam) 700 hPa heights & anomalies (dam)

700 hPa heights & anomalies (dam) 1 day before onset of heat wave 700 hPa heights & anomalies (dam) 1 day before onset of heat wave 700 hPa height anomaly correlations (from Namias 1955)

Loop 2: Synoptic Evolution of Heat Waves

Synoptic Evolution of Heat Waves (4 Days Prior) n= hPa heights (dam), & winds (m s -1 ) 700 hPa heights (dam), Temp. (°C, red), & winds (m s -1 )

Synoptic Evolution of Heat Waves (2 Days Prior) n= hPa heights (dam), & winds (m s -1 ) 700 hPa heights (dam), Temp. (°C, red), & winds (m s -1 )

Synoptic Evolution of Heat Waves (First Day) n= hPa heights (dam), & winds (m s -1 ) 700 hPa heights (dam), Temp. (°C, red), & winds (m s -1 )

Synoptic Evolution of Heat Waves (Midpoint) n= hPa heights (dam), & winds (m s -1 ) 700 hPa heights (dam), Temp. (°C, red), & winds (m s -1 )

Synoptic Evolution of Heat Waves (Last Day) n= hPa heights (dam), & winds (m s -1 ) 700 hPa heights (dam), Temp. (°C, red), & winds (m s -1 )

Loop 3: Air Mass Transport

Synoptic Evolution of Heat Waves (4 Days Prior) n= hPa: 18°C isotherm

Synoptic Evolution of Heat Waves (2 Days Prior) n= hPa: 18°C isotherm

Synoptic Evolution of Heat Waves (First Day) n= hPa: 18°C isotherm

Synoptic Evolution of Heat Waves (Midpoint) n= hPa: 18°C isotherm

Synoptic Evolution of Heat Waves (Last Day) n= hPa: 18°C isotherm

Conclusions

Conclusions – Statistical Analysis Negative trend in Northeast & Southeast hot days and heat waves turns positiveNegative trend in Northeast & Southeast hot days and heat waves turns positive 1950’s, 1980’s, and 1990’s are peak in heat wave frequency1950’s, 1980’s, and 1990’s are peak in heat wave frequency Although weaker, trends seen in other eastern regions except SouthernAlthough weaker, trends seen in other eastern regions except Southern Trends east of Rockies match those found by DeGaetano and Allen (2002); Gaffen and Ross (1999)Trends east of Rockies match those found by DeGaetano and Allen (2002); Gaffen and Ross (1999)

Conclusions – Antecedent Conditions Correlation between strong 700 hPa anticyclone in eastern Pacific and positive height anomalies over Northeast U.S. in agreement with past teleconnection researchCorrelation between strong 700 hPa anticyclone in eastern Pacific and positive height anomalies over Northeast U.S. in agreement with past teleconnection research Enhancement and eastward progression of ridge over North America coupled to amplification of ridge (trough) over Eastern Pacific (West Coast)Enhancement and eastward progression of ridge over North America coupled to amplification of ridge (trough) over Eastern Pacific (West Coast) Origin of wave energy appears to be equatorial PacificOrigin of wave energy appears to be equatorial Pacific

Conclusions – Heat Wave Evolution “Kicking trough” builds ridge in the East, displacing warm air mass eastward“Kicking trough” builds ridge in the East, displacing warm air mass eastward Northeast located underneath both anticyclonic shear side of jet exit and downstream of ridge axis (favored area for subsidence)Northeast located underneath both anticyclonic shear side of jet exit and downstream of ridge axis (favored area for subsidence) Westward extension and intensification of Bermuda High allows for west-northwesterly (downsloping) flow into NortheastWestward extension and intensification of Bermuda High allows for west-northwesterly (downsloping) flow into Northeast

Statistical Analysis (cont.)

14 stations

54 stations