Sie sind auf Seite 1von 91

c   


  

  


Paul Kocin

National Press Foundation


Norman, Oklahoma
March 14, 2007
AGENDA
ð1) Winter Weather is Pretty
How serious is winter weather?
So why should journalists
care?
ð2) Winter forecast issues
You should know

ð ) How have winters been


affected By Global
Warming

ð4) Northeast Snowfall Impact Scale


ðNESIS)
ð1) Winter Weather is Pretty
How serious is winter weather?
So why should journalists care?
shis is why
 Each year, automobile accidents claim an
incredible amount of lives, cause injuries
and property damage.
she average/year:
² 41,000 deaths,
² million injuries,
² billions in damage.
 Most of the above is NOs weather
related, BUs about 15% is!
 Although 15% sounds like a
small percentage, that still
leaves approximately
 7,000 deaths
 00,000 injuries
 and billions of dollars in
damage per year!
 et·s put these numbers in perspective:

Deaths ð10-year average)

Hurricanes are responsible for about


1 deaths /yr ðthat will change)
sornadoes are responsible for about
5 deaths /yr
Floods are responsible for about
4 deaths /yr
Heat Waves are responsible for
about 2 7 deaths /yr
 Adverse weather is
associated with over 1.5
million crashes per year.

 she adverse weather is


typically rain, snow/ice and
fog.
 Examples: In 2001
² 5400 deaths linked to rain-slicked
roads
² 1100 deaths linked to snow
and ice
² 670 deaths linked to fog

 ost estimates
² Injuries, loss of life, and property
damage cost an average of 42
billion dollars annually just for
snow and ice!
Es·S PUs A sHIS IN PERSPEsIVE:
‡ Most of us are more likely to be
affected through injury, death, or
property damage from road
weather than from more well-
known weather issues such as
hurricanes, tornadoes, and flash
floods combined.
AND FINAY:
‡ And snow and ice make up a
significant component of the
dangers posed to transportation
nationwide
So why should journalists care?
‡ While ´eventsµ make the ´newsµ,
there are less sexy weather
hazards that impact far more
people
Agenda

ð2) Winter forecast issues


You should know
Winter vs Severe Storm vs Hurricane
forecasting
very different kinds of problems

0 Hurricanes involve assessing track, intensity


and size of circulation, storm surge, flooding
rains and timing

0 Severe Storms involve pinpointing areas


when and where tornadoes, hail and strong winds
will occur

0 Winter storms involve determining


precipitation type, amount, location and timing
Winter forecast issues
O she main issue: Rain vs
freezing rain vs sleet vs snow

O Predictability & Uncertainty

O Winter climatology-means and


extremes ² ðhow common vs how
rare)
So how do we forecast a winter
storm?
O Depends on how far out from
the event
O ong range and short range
models
O Data!
When do we see something coming?
O Right now, there are patterns that
sometimes foretell that winter
storms may be affecting our area

O It·s often not El Nino ðalthough


sometimes that has a big role)
O NAO
When do we see something coming?

O Sometimes we see signals that suggest


a major snow is possible as much as 10
days in advance

O a lot of these signals are ´false alarmsµ

O As we get to to 5 days in advance, we


start getting excited ðso do you!)
she tools we use

O A magic wand?
O she farmers
almanac?
O Uncle Fred·s
bunions?

O Models«««
What are models?
O Very complex computer programs based
on basic physics and physical
parameterizations run on some of the
world·s most powerful computers

O these models compute trillions of


numbers that predict the future state of
the atmosphere and can be plotted on
maps

O meteorologists love to look at model


data on maps!!!!!
urrent models
 ong range
she GFS ² up to 15 days

 Short range
NAM & others ðMM5, RU, GFS good
too)
every hrs thru 4 hours ð 1/2 days)

 Foreign models
EMWF
UKMEs
anadian Model

and there are others, ensembles,


etc««.
When the %!$#!! is hitting
the fan
 Radar ² much better with snow than
only 15 years ago

 Surface weather charts


My ´secret weaponµ ² try to plot
every hour
allows me to see if storm is going
according to plan or not

 Satellite data, model updates, coop


observers
webcams, colleague consultation
Winter forecast issues

O During winter, people expect us to


accurately predict precipitation amounts ðin
terms of snow depth) ² need to express
´uncertaintyµ

O surns out that we have little skill in this


area, especially more than 6 hours out

- few mind if 1 inch of rain versus .25 inches


falls
SNOWFA MEASUREMENsS

O snowfall is surprisingly tough to measure


´accuratelyÒ

O Unfortunately, the ´officialµ measurements


often do not capture real variations in
snowfall that can and do occur over small
distances
Measuring snow

 Snowfall vs snow depth

 Measuring at differing intervals may


get you different amounts
 Snow melts, compacts, sublimates,
drifts, refreezes
 It accumulates differently on
different surfaces
 It·s a messy substance to measure

 sherefore, it·s even more messy to


forecast
Winter forecast Reality

O Precipitation ðand that means snow)


remains a meteorological issue that is still
not well resolved

O Meteorologists have been providing


forecasts of precipitation amounts for years
ðin terms of snowfall forecasts)

it has almost been a crapshoot ²


O does it make a difference whether it is a 1-
day, 2-day, - day or 7-day forecast!
Winter forecast issue Reality
O shat said, precipitation forecasting has
improved to a point where forecasts are
skillful 24 to 4 hours out

ðthere are, of course, exceptions!!)


Winter forecast issue Reality
O Not all weather situations are ´equally
predictableµ
O And forecasters don·t necessarily know
when a situation is more predictable than not
ðwe usually find out when you do - after the
stuff happens)
Winter forecast issue Reality
O Sometimes all the models are in
agreement!
O Problem is, model agreement doesn·t mean
accurate forecast ² that·s where the science
is
she main issue: Rain vs
freezing rain vs sleet vs snow
„ Winter storms are usually more complicated
than this!!!!!!!
s  pr„lm wit f„casting avy
sn„
 It often occurs close to the
rain/snow line

 It sometimes occurs close to the


northwestern edge of the snow

 50-mile errors in location produce


big problems!

 she ´dreaded dry slotµ

 Forecast amounts depend on how


it·s measured and measuring
snow is not as easy as it sounds
Example
ook at this narrow band of snow!
Another example
One of the craziest snows in recent memory!!
÷ t f„t  
Predictability

Some storms are easy«««.

ŏ Affect one small region ² it·s well


forecast
ife is GREAs

ŏ Most storms are widespread, multi-


day, multi-form events««««..
forecasts can be good one
place, lousy others««..lots of areas
for
major errors!
More predictability

0 Whether big or small, winter storm


predictability is very variable!!!

0 Some winter storms are amenable to


prediction ² even several days in
advance

0 Other winter storms seem to be


unpredictable even up to the day of
the event
Uncertainty

0 shere is always uncertainty when


predicting snowfall amounts

0 Uncertainty can be expressed


probabilistically

0 Uncertainly doesn·t mean ´we don·t


knowµ, but represents a continuum
from what we do know versus what
we don·t
Winter climatology-means and extremes ²
ðhow common vs how rare)

Mean Annual Snowfall

„ m
„t„
Ò
„
Ò
„ „ 

Ò „
„ „ „ Y„
 ÷ t„
 „ Ò
„
j Ò
„
Al
Ò „ Atlt
„ Ò
ll

 m
„
÷ t l mt„l„
÷ t l mt„l„
A

ð ) How have winters


been
Affected By Global
Warming?
How cold was this winter?
ðDecember thru February)

45th warmest in 112 seasons


How did it break down
month to month?
Past winters: Warm or cold?
Nationwide Ranking
2006/2007 45th Warmest
2005/2006 rd Warmest

2004/2005 10th Warmest


200 /2004 42nd Warmest
2002/200 1st Warmest
2001/2002 5th Warmest
2000/2001 26th oldest
1999/2000 Warmest
No shortage of major winter storms
During the same period
January 25, 2000 January 27-2 , 2004
December 0, 2000 January 22-2 , 2005
February 5, 2001 December 9, 2005
March 5-7, 2001 February 11-12, 2006
December 25, 2002 February 1 -14, 2007
January -4, 200
February 15-17, 200
December 5-7, 200

And these are just major Northeast


Winter Storms
Regional semperature omparison: 6 previous winters

2005/
2006 2002/
200

2004/
2005 2001/
2002

200 /
2004 2000/
2001
÷  t·  m, t·  m
A  tt

„ ml J  

 f
÷mt J 
„ „
 „ t 
„t tl
A t   t   m l

m 

t mt
m „ „
 t  „   „l „t

  

t „lt  


So, how have winters
been doing in the Global Warming era?

„ „t   m  · j



mt  „   lt 8  t

„  tt m  m  t t


 t „t „t     t „l  „

„ t ll  t „ „m   l 


 t t„m  t mt
 t„m lt 
 t.)
We·ve come up with a scale
that estimates winter storm impact
For Northeast Winter Storms

„      t„ t  lt t„  „


t  tt „ ..
A

ð4) Northeast Snowfall Impact Scale


ðNESIS)
Fujita and Saffir-Simpson Scales

‡ Measure potential for


ss !
to property and
loss of life from
tornadoes and hurricanes

‡ ategories (0 or 1 thru )

‡ Scale used during (Saffir-


Simpson) or after (Fujita)
the event
S S

‡ Measures potential for human and economic


 S"s !

‡ ategories (1 thru )

‡ se during and after


storm
S!WST!M M"AT  T!

‡ S!WFA
‡ W  S"
‡ TM"AT
‡ AT !
‡ TS TY
‡ ST!M ´ TS TYµ
‡ G! TM"ATS
‡ S AG
‡ T M !F AY, T., T. T.
‡ T#
T!! !M" AT#####
‡ !!$ AT 30 ASS
² 19 0-2000

‡ !T! S!WFA
at 4µ, 10µ and at Fe 198
10µ intervals

‡ TH!S !T!S
"ST:
(1) an AA Mar 1993
(2) an AFFT "!"AT !
What S S is

‡ ß   
  
        
  

‡       


      
How to get estimates of area and population?
‡ G S Systems

F 1978 MA 1993

‡ nitially done for 30 cases for 4µ, 10µ, 20µ.. ..tc.


‡ A mean Area and mean "opulation were computed for 30
cases
R  
    R 
      
SM (AAS  "!"AT !S)
!TA   AH S!WFA T%A

A= Area of snowfall within a given contour interval


Amean = mean area of 30 cases

"= "opulation of area within a given contour interval


"mean = mean population for 30 cases

n = nterval number (>4µ;=0.4; 10µ; =1; >20µ, =2; >30µ, =3 ..)

And applied the equation to all 30 cases


H„  S S l ?

‡ Al  t„ t  t„tl „ll  t  t „


„t t   t „ R„ )
xamples
Mostly 4-10µ, small 10µ

NESIS VAUES
lose to 1
arger area of mostly 4-10µ

NESIS VAUES
lose to 2
Small 4µ, arger area of 10µ, small 20µ

NESIS VAUES
lose to
arger 4µ, arger area of 10µ, small 20µ

4µ 10µ NESIS VAUES


20µ Greater than 4
ven larger area of 4µ, 10µ, 20µ

NESIS VAUES
Greater than !
Mean = 4.80
Tested the scale on 70 cases:

‡ 30 Major Snowstorms
‡ 1 ´ nteriorµ
Snowstorms
‡ 1 ´Moderateµ
Snowstorms
‡ 4 Historic Snowstorms
‡ ecent Snowstorms
incl: Winter of
2002/2003
‡ Then came up with
ategories
S S ATG! S
ATG!Y S S %AS  „ ASS S "T !

‡ j j ² . &&
'!TAÒ
‡ . 
.&& 'S G  ATÒ
‡
² .&& j 'AJ!Ò
‡   &.&&  ' "" GÒ
‡ j .  '(TÒ
she Blizzard of ¶05

NESIS~6.10
As 4
she Blizzard of ¶06

 ).j
As
she Valentine·s Day Storm of 2007

 ).j
As
ortheast Snowfall mpact
Scale
Top anked Snowfalls for
70 major snows
ank AT S S ATG!*
‡ j j j A j&&
j .
‡ 8 JA j&& jj.
‡
j j8  00
8.&j
‡ jjj A j888 8.

‡ jjj  j8&& 8.jj
SMMA*

‡ S S is an !bjective Scale that measures the impact of


ortheast snowstorms on the densely populated rban
orridor
‡ A Scale erived from 30 Historical ases (19 0-2000)
‡ omputed %alues mapped to Specific ategories (1- )
SMMA* (ont·d)

‡ Application appears to separate the most memorable


storms from others with subjectively smaller impacts
‡ The scale provides an objective number for total snowfall
distribution and relates impact that matches a
subjective view of the snowfall maps
sH 
G„ t„m  #
÷ t l mt„l„

„·t
÷ t l mt„l„

l 
÷ t l mt„l„

 l H„„
÷ t l mt„l„

‡ t„m  lt „


 A l „ll

Ò
Ò

Ò
Ò

j Ò Ò
Ò
÷ t l mt„l„
÷ t l mt„l„
÷ t l mt„l„
÷ t l mt„l„

Das könnte Ihnen auch gefallen