Outline:

1
Outline

• What do current weather forecasting skills tell
  us about the state of the art of the science and
  operations?
• Why do surface boundaries driven by differential
  diabatic heating and cooling often cause forecast
  problems?
• The 500 hPa height forecast was nearly "perfect",
  so why isn't everyone cheering?
• What are some of the uncertainties and challenges
  associated with the forecasting of TCs and MCSs?
• Why should you care about phase space,
  probability density functions and ensemble
  forecasting?
• What science/forecast opportunities arise when
  mesoscale disturbances can be represented on
  upper-level analyses?

2
Some Measures of Forecast Skill
3
(No Transcript)
4
(No Transcript)
5
Day at which fcst loses useful skill (AC0.6) N.
Hemisphere calendar year means
Forecast day
6
Day 7
Day 5
Day 3
In last 10 years, HPC forecasts have added 2
days of skill
7
National Weather Service Probability of
Precipitation Reliability October 1994 to
September 2004
40,672
-
N 2,225,086
-
30,377
Observed Frequency ()
-
47,142
Number of Forecasts
-
60,369
-
82,664
-
131,992
-
198,343
-
235,835
-
307,835
-
121,819
-
878,764
Forecast Frequency ()
8
(No Transcript)
9
Source http//www.hpc.ncep.noaa.gov/research/ams
ver/sld007.htm
10
Source http//www.hpc.ncep.noaa.gov/html/hpcverif
.shtml
11
Source http//www.hpc.ncep.noaa.gov/html/hpcverif
.shtml
12
(No Transcript)
13
Source http//www.hpc.ncep.noaa.gov/research/ams
ver/sld011.htm
14
Source http//www.hpc.ncep.noaa.gov/research/ams
ver/sld012.htm
15
Source http//www.hpc.ncep.noaa.gov/research/ams
ver/sld013.htm
16

• Consider Equitable
• Threat Scores
• ETS location-dependent,
• related to climatological
• probability.
• (2) Average of ETS at
• individual grid points 0.28
• (3) ETS after data lumped into
• one big table 0.42

Source Hamill and Juras (2007), Measuring
forecast skill Is it real skill or is it the
varying climatology?
17
Brier Score (BS) and Brier Skill Score (BSS)
0 BS 1
N number of forecast-event pairs F
forecast probability 0 observed value
(0 or 1) BSS 1 - BS/BSref BSref
reference score (e.g. climatology or consensus)
18
(No Transcript)
19
Source www.cpc.ncep.noaa.gov
20
(No Transcript)
21
Washington Post 2 March 2005
22
Questions that need to be asked and answered
before making a weather forecast
23
Forecasting Philosophy Bosart (2003)
What has happened? Why did it happen? What is
happening now? Why is it happening? What is
going to happen? Why is it going to happen?
24
Outline

• What do current weather forecasting skills tell
  us about the state of the art of the science and
  operations?
• Why do surface boundaries driven by differential
  diabatic heating and cooling often cause forecast
  problems?
• The 500 hPa height forecast was nearly "perfect",
  so why isn't everyone cheering?
• What are some of the uncertainties and challenges
  associated with the forecasting of TCs and MCSs?
• Why should you care about phase space,
  probability density functions and ensemble
  forecasting?
• What science/forecast opportunities arise when
  mesoscale disturbances can be represented on
  upper-level analyses?

25
The Northern California Cold-Air Damming Event
of 25 January 2008
26
Source http//www.maps-for-free.com/index.html
27
Source http//www.maps-for-free.com/index.html
28
0643 UTC 25 Jan '08
Source http//www.maps-for-free.com/index.html
29
1243 UTC 25 Jan '08
Source http//www.maps-for-free.com/index.html
30
Source http//www.maps-for-free.com/index.html
31
Source http//www.maps-for-free.com/index.html
32
The Southern New England Coastal Front and Heavy
Snowband Eventof 27-28 January 2008
33
(No Transcript)
34
DT pressure (hPa), 850-DT shear (wind barbs in
kt) and 925-850 mean ? (every 0.5x10-4 s-1)
Source www.atmos.albany.edu/facstaff/rmctc/DTmaps
/animSelect.php
(hPa)
35
AREA FORECAST DISCUSSION NATIONAL WEATHER SERVICE
TAUNTON MA 700 PM EST SAT JAN 26
2008 ..SYNOPSIS... LOW PRESSURE WILL DEVELOP OFF
THE MID ATLANTIC COAST...THEN WILL
STALLSOUTHEAST OF NEW ENGLAND AS IT STRENGTHENS
INTO A LARGE OCEAN STORM LATE SUNDAY THROUGH
MONDAY NIGHT. THIS STORMWILL LIKELY BRING A
MYRIADOF HAZARDS TO THE CAPE AND ISLANDS.
SHORT TERM /6 AM SUNDAY MORNING THROUGH 6 PM
SUNDAY/... SUNDAY...LOW LEVEL MOISTURE EXPECTED
TO LINGER ACROSS THE FORECAST AREA MUCH OF THE
DAY GIVEN INCREASING ONSHORE FLOW. QUESTIONS
REMAIN REGARDING THE AMOUNT OF FOCUS FOR LIFT
WHICH OCCURS.
36
AREA FORECAST DISCUSSION NATIONAL WEATHER SERVICE
TAUNTON MA 1020 AM EST SUN JAN 27
2008 ..SYNOPSIS... LOW PRESSURE OFF THE MID
ATLANTIC COAST EARLY THIS MORNING WILL RAPIDLY
INTENSIFY INTO A STRONG STORM TODAY. WIND GUSTS
UP TO 60 KNOTS WILL OCCUR NORTH AND WEST OF THE
STORM CENTER AS IT PUSHES TO ABOUT 150 NM S OF
NANTUCKET THIS EVENING. .NEAR TERM /UNTIL 6 PM
THIS EVENING/... SNOW SHOWERS HAVE INCREASED IN
AREAL COVERAGE ACROSS EASTERN MASSACHUSETTS IN
RESPONSE TO A WEAK UPPER LEVEL TROUGH LOW
LEVEL MOISTURE CONVERGENCE (COASTAL FRONT) AND
LOW LEVEL INSTABILITY OFF THE WARMER OCEAN
WATERS. THIS BATCH IS EXPECTED TO ROTATE TO THE
WEST AND DIMINISH WITH TIME GIVEN THE CHANGE IN
LOW LEVEL WIND TRAJECTORY TO THE EAST THEN
NORTHEAST. GIVEN THE LATEST REPORTS CALLED IN
TO ESSEX COUNTY...HAVE DECIDED TO UPSHIFT TO A
SNOW ADVISORY. RECEIVED GENERALLY 1 TO 2 INCH
REPORTS ACROSS ESSEX COUNTY EXCEPT 3 TO 4 INCHES
IN THE IMMEDIATE GLOUCESTER AREA.
37
(No Transcript)
38
(No Transcript)
39
AREA FORECAST DISCUSSION, TAUNTON MA (2132 UTC 27
JAN08) .NEAR TERM /UNTIL 6 AM MONDAY
MORNING/... LOW PRESSURE CONTINUES TO TRACK WELL
SOUTHEAST OF NANTUCKET ISLAND.THE LATEST
NAM/UKMET AND CANADIAN RGEM HAVE TRENDED FURTHER
EAST WITH THE TRACK OF THE LOW...INDICATIVE OF A
MORE PROGRESSIVE SCENARIO AND THEREFORE LOWER
QUANTITATIVE PRECIPITATION FORECAST AMOUNTS
EXPECTED. OVERALL HAVE DECIDED TO CONVERT WINTER
STORM WARNINGS TO SNOW ADVISORIES ACROSS CAPE COD
AND THE ISLANDS AS WELL AS COASTAL PLYMOUTH
COUNTY. AREA FORECAST
DISCUSSION, TAUNTON MA (2244 UTC 27JAN08) NEAR
TERM/UNTIL 6 AM MONDAY MORNING/... WORST OF THE
STORM WILL OCCUR NEXT 1-2 HOURS OVER THE CAPE.
DONT THINK HEAVY SNOW ON THE CAPE LASTS MUCH
PAST 7 PM. BAND WILLLIKELY WEAKEN AND CONTRACT
AS IT MOVES WEST OF THE CANAL BUT A BRIEF PERIOD
OF MODERATE SNOW POSSIBLE SE MA THIS EVENING.
40
AREA FORECAST DISCUSSION, TAUNTON MA (0134 UTC 28
JAN08) .NEAR TERM /UNTIL 6 AM MONDAY
MORNING/... WE COULD NOT SIT ON THIS WITHOUT A
WARNING. CERTAINLY NEAR BLIZZARD CONDITIONS AT
THIS TIME AND BIG SNOW GROWTHPOSSIBLE IN BANDS
THRU 06Z. TOO MUCH 1/2 SM IN SNOWDRIFTS
GREATER THAN A FOOT IN SE PYM COUNTY. WE ALSO
HEAR OF 2 INCHES/HR VCNTY SANDWICH ENDING 730PM.
6 INCHES AT SANDWICH. BREWSTER 7 INCHES
NOW. AREA FORECAST
DISCUSSION, TAUNTON MA (0242 UTC 28JAN08) .NEAR
TERM /UNTIL 6 AM MONDAY MORNING/... MAJOR
STORM IN PROGRESS IN SE NEW ENGLAND DONT
TRAVEL SRN PLYMOUTH COUNTY TO CAPE COD IF YOU
DONT HAVE TO. THIS IS A PBLM STORM. I SEE 16
INCH MAX SNOWFALL IFMEASURING CAN BE DONE
ACCURATELY IN THESE CONDS.
41
Source http//www.rap.ucar.edu/weather/radar/
42
IR and WV Satellite Imagery for 0245 UTC 28 Jan'08
WV
IR
Source http//www.rap.ucar.edu/weather/satellite/
43
MM5 Forecasts from SUNY/Stony Brook
3 h Fcst VT 0300 UTC 28 Jan'08
6 h Fcst VT 0600 UTC 28 Jan'08
700 hPa Z/?/RH
Precip last 3 h (0.01")
Source http//cheget.msrc.sunysb.edu/html/alt_mm5
.cgi
44
MM5 Forecasts from SUNY/Stony Brook
3 h Fcst VT 0300 UTC 28 Jan'08
6 h Fcst VT 0600 UTC 28 Jan'08
925 hPa winds/700 hPa ?
700 hPa Frontogenesis/Base Reflectivity
Source http//cheget.msrc.sunysb.edu/html/alt_mm5
.cgi
45
Outline

• What do current weather forecasting skills tell
  us about the state of the art of the science and
  operations?
• Why do surface boundaries driven by differential
  diabatic heating and cooling often cause forecast
  problems?
• The 500 hPa height forecast was nearly "perfect",
  so why isn't everyone cheering?
• What are some of the uncertainties and challenges
  associated with the forecasting of TCs and MCSs?
• Why should you care about phase space,
  probability density functions and ensemble
  forecasting?
• What science/forecast opportunities arise when
  mesoscale disturbances can be represented on
  upper-level analyses?

46
A predecessor heavy rain event in Hurricane
Frances on 8 September 2004
47
Frances storm track
48
Frances storm precip
49
1200 UTC 8 Sep'04
50
(No Transcript)
51
FWN
NYC
JFK
09Z/08
12Z/08
100
KOKX
200
300
500 J/kg
500
700
15Z/08
12Z/08
850
1000
52
Surface ______ SLP (hPa) ______ ? (K)
1200 UTC 8 Sept 04
53
Meteograms
JFK
1018
1017
Legend T, Td(C) Hourly rain (mm)
NYC
1018
SLP (hPa)
1017
1016
FWN
1019
1018
09
10
11
12
13
14
15
Time (UTC) 8 September 2004
54
E-W (41.5N) ? (K), winds (kt), and vertical
motion (shaded, 10-3 hPa s-1) 1200 UTC 8 Sep'04
55
Outline

• What do current weather forecasting skills tell
  us about the state of the art of the science and
  operations?
• Why do surface boundaries driven by differential
  diabatic heating and cooling often cause forecast
  problems?
• The 500 hPa height forecast was nearly "perfect",
  so why isn't everyone cheering?
• What are some of the uncertainties and challenges
  associated with the forecasting of TCs and MCSs?
• Why should you care about phase space,
  probability density functions and ensemble
  forecasting?
• What science/forecast opportunities arise when
  mesoscale disturbances can be represented on
  upper-level analyses?

56
Personal Phase Space
Good
Bad
None
Caffeine
Lots
57
Forecast Model Climatological Phase
Space (Ensemble Forecasting with One Model)
Wet
Model 1
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
Rainfall
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
Dry
Cold
Temperature
Warm
58
Forecast Multimodel Ensemble Climatological Phase
Space
Wet
Model 2
Model 3
Rainfall
Dry
Cold
Temperature
Warm
59
(No Transcript)
60
(No Transcript)
61
Climate Change Scenarios
?T2
?T1

62
24 h Forecast Area gt 2.5 cm Precipitation 1970
63
24 h Forecast Area gt 2.5 cm Precipitation 1990
64
24 h Forecast Area gt 2.5 cm Precipitation 2008
Observed
Model 1
Model 2
Model 3
Model 4
Basis for Ensemble Probability Forecasting
65
Penn State GEFS F-108 h SLP and 1000-500 hPa
Thickness VT 1800 UTC 8 Mar'08
Penn State GEFS F-108 h 500 hPa
Heights VT 1800 UTC 8 Mar'08
http//www.meteo.psu.edu/gadomski/ewallmref.html
66
Mike Bodner NCEP/HPC Realtime Standardized
Anomalies
Source http//www.hpc.ncep.noaa.gov/training/SDs/
67
Outline

• What do current weather forecasting skills tell
  us about the state of the art of the science and
  operations?
• Why do surface boundaries driven by differential
  diabatic heating and cooling often cause forecast
  problems?
• The 500 hPa height forecast was nearly "perfect",
  so why isn't everyone cheering?
• What are some of the uncertainties and challenges
  associated with the forecasting of TCs and MCSs?
• Why should you care about phase space,
  probability density functions and ensemble
  forecasting?
• What science/forecast opportunities arise when
  mesoscale disturbances can be represented on
  upper-level analyses?