Time and location:
Fall 2003 - Tues and Thurs:
Instructor:
J. M. Fritsch
603
863-1842
fritsch@ems.psu.edu
Office hours:
Teaching Assistant:
Matt Greenstein
mdg197@psu.edu
This is an information access site for Meteo. 415, Forecasting Practicum.The information consists mainly of:
Philosophy of Course
Meteo 415 is a forecasting practicum that provides an
opportunity for students to obtain real-time real-world experience forecasting
conventional weather parameters at selected cities in the
Grades
Grades will be based on two real-time forecast contests (25% each), class assignments and discussion (25%), and a term project (25%). Any person who has taken Meteo 415 previously must register for Meteo 496 (Independent Studies) instead of Meteo 415. Grading of repeaters will be done separately from first time students.
Academic
Integrity: This course follows the
Term Projects
Term projects may be on any topic having to do with weather forecasting. For example, the project may consist of developing a forecasting technique for a given weather condition that interests you (e.g., severe downslope windstorms) or for a given location for which there is no conventional guidance available (e.g., a ski resort). Or, it may be a term paper to find out how to do forecasts for special sectors of the economy. For example, if you are interested in aviation forecasting, you may want to find out what products are most helpful to commercial pilots? What techniques are applied at commercial airlines? What forecasts matter most to the utility industry? How are they produced? How could they be improved? What are the special weather forecasts that serve the agriculture industry? The insurance industry? How can weather risk be quantified? Another project option would be to compare various operational products such as the new ETA MOS versus the AVN MOS. Or, you could compare the performance of individual TV station forecasts to the National Weather Service forecasts to see if the TV forecaster is actually better (as they so often claim)! You could research how ensembles are created and how best to use them? In short, this is an opportunity to explore just about any aspect of forecasting that has intrigued you. Since the project counts as much as one of the forecasting contests, it is to the student’s advantage to select a topic ASAP and get it approved by the instructor. To receive approval, you must submit a title, objective, methodology, and an outline of the major sections of the project at any time prior to, but no later than, 25 September.
Post-mortems
Students will be assigned to perform post-mortems of the forecasts made in class. Teams composed of two or three class members will be constructed at the start of the semester. A schedule for when each team performs a post-mortem will be posted on this web site. The team post-mortems will always be done on Thursdays (at which time the forecasts made on Tuesdays will be reviewed).
The following factors should be
discussed in the post-mortems:
· Primary weather threat.
· Synoptic/mesoscale
situation.
· Guidance.
· Team member forecasts.
· Verification.
Post-mortems of forecasts made on a Thursday will be done the following Tuesday. These post-mortems will be conducted by the two individuals who receive the lowest scores on the forecast made on the previous Thursday. Once the Thursday forecast is verified and scored, these individuals will be notified by email of their obligation for the following Tuesday.
Assignments
Unless specified otherwise, assignments are due one week from the date on which they were assigned.
Class Notes
It
is recommended that you purchase the class notes (available at the Penn State
Bookstore). The notes contain a great deal of helpful information (e.g., how to
use MOS and FOUS, descriptions of the numerical models, how to forecast
temperature, wind, precipitation type, etc.). It is to your advantage to
have them available.
Remarks
Forecasting
is serious business. While most of us are here because we enjoy the weather and
the challenge to forecast it, we must remember that forecasting carries with it
a large responsibility. Based upon our
forecasts, people make decisions. If we are wrong, most of the time it
results in only a minor inconvenience to most people. However, there are those
portions of the business, transportation, military, industrial and recreational
communities that are seriously impacted by weather changes - - - and busted
forecasts mean significant revenue losses. For example, an error of just
a few hours in the forecast time that ceiling and/or visibility will fall below
the minimum legal operational levels at a single airport can cost commercial
airlines a quarter of a million dollars a day. Similarly, errors of
only a few °F in the maximum temperature forecast can, on extremely hot days,
result in losses of millions of dollars to certain sectors of the utility
industry. Of course, the ultimate consideration is people's lives - - -
here the burden is heaviest. Too many times forecasters have become trapped in
a daily routine, unable or unwilling to recognize those situations that depart
radically from the norm. It is extremely difficult for forecasters to
recognize when a life-threatening but truly rare event is upon them - -
- it is even more difficult to take appropriate action. For
example, consider the situation that occurred in northern Utah in August of
1999. Conditions became favorable for tornadic
storms one afternoon. But, since such storms virtually never occur in
this region, especially not in August, forecasters were reluctant to issue the
appropriate warnings for the tornado that swept through downtown Salt Lake
City. Weather forecasters are not alone in being reluctant to act when faced
with such rare events. As a matter of course, humans in general
have great difficulty abandoning the stereotype and taking a new and unusual
course of action in response to conditions that are very different from what
they are used to, even when, intellectually, they know that they should. This
human characteristic is commonly called "mindset" and is one reason
why the National Weather Service has a Storm Prediction Center in Norman,
Oklahoma and a Heavy Precipitation Forecasting Branch at the National Centers
for Environmental Prediction in Washington, D.C. These operations
are geared to look for the unusual -- and are less likely to be mesmerized by
the multitude of mundane events that forecasters at regular NWS offices or
private-sector operational centers encounter over the years. Students,
however, are unencumbered by years of the mundane. As a result, they
typically have the opposite problem -- they try to turn routine events into the
unusual! One of their most common problems is that they "overforecast".
For example, a one to three inch snow is forecast to be four to six
inches. Since the definition of "heavy" snow is four inches or
more, the overforecast means that heavy snow warnings
must be issued and this causes many actions (or cancellations of actions), some
of which can be very expensive, to be taken by the general public.
Similarly, students tend to forecast record-breaking events more frequently
than the conditions warrant. Therefore, when you prepare your forecasts this semester,
try to keep this tendency to overforecast in
mind. Be conservative. In the long run you will score better and
develop a much greater credibility with your family, friends and peer
professionals. It might help your M415 grade too!
9/4
- 9/11 Course Information
- term project
- available data: local sources; web sites
- Game rules
- scoring (Brier score)
- practice forecasting
9/18 Game I begins
9/25 Term project outline due:
Include title, objective, methodology,
and an
outline of the major sections.
10/10 Fall Break
10/28 Last day of Game I
10/30 Game II begins
11/27 Thanksgiving Break
12/11 Game II ends
12/15 Term projects due
-
- - - Topics for In-class Discussion - - - -
- MOS and FOUS
General
approach to making a forecast
- who is your client
- climatology
- persistence
- extrapolation
-
pattern recognition
- local effects
- decision trees
- cost-benefit analysis
Numerical
models
- sensitivity to initial conditions
- consistency
- systematic errors
- deficiencies in model physics
- frontal analyses
- model consensus
Temperature
- free-air extrapolation and adjustment
- analogous thickness/pressure pattern
Winds
- geostrophic
- analogous patterns
- diurnal cycle (effect of mixing, clouds,etc.)
Icing
Fog
The future of weather forecasting
Description and Scoring
Students are provided with a spreadsheet of weather parameters to be forecast. The parameters are organized into the following groups:
1. temperature,
2. precipitation,
3. type, and
4. events.
Each group contains four or more weather parameters. The parameters are selected based upon the current weather situation. Typically, however, many of the parameters will be the same from one forecast day to the next. For example, the “temperature” group will usually include the maximum and minimum temperatures in selected time periods. The “precipitation” group will normally require a forecast of the probability of measurable (0.01 inches) precipitation in a specified time period, as well as the amount and duration of the precipitation. Dewpoint, cloud cover, wind speed, and wind direction are usually included in the “type” group, while parameters such as snow, freezing rain, thunderstorms, fog, ceiling, and visibility, normally appear in the “events” group.
On a typical day, forecasts are made for two or three cities. Occasionally, if we are pressed for time, forecasts are made for only one city.
All forecasts are probabilistic and all scoring is performed using the Brier Score (SB). The Brier score varies from zero (best forecast) to one (worst forecast) and is computed in the following manner:
SB = (F-O)2
Where F is the forecasted probability (expressed as a decimal between zero and one) and O is the observed outcome of events. The observed outcome will be either 0 (it didn’t happen) or 1 (it did happen). For example, here are some sample Brier Score calculations for forecasts of the probability of precipitation (pop):
Forecast Observed Error Calculation
pop:
60%
100% (0.6 - 1.0)2
= 0.16
pop:
90%
100% (0.9 - 1.0)2
= 0.01
pop:
60%
0% (0.6
- 0.0)2 = 0.36
pop:
90%
0% (0.9
- 0.0)2 = 0.81
The Overall score for a given forecast is the sum of
the error points for all of the parameters in all of the weather groups for all
of the cities. Absentee forecasters will be assigned the mean class score for
each day that is missed. After the scores are compiled, they are
ranked. After each forecast, a mean ranking is computed for each student
and the mean rankings are then ranked. Ranking ensures that each forecast
day is weighted the same as any other forecast day, i.e., days with
exceptionally high or low scores (because the forecast happened to be very easy
or very difficult) can not skew the results of the competition. The
competition is divided into two contests so that students can have a second
opportunity to demonstrate their skills without being encumbered by an unlucky
or unrepresentative forecast.
^ Introduction
Rules and General Information:
The valid time or time period for which each weather parameter is being forecast will be listed (on the spreadsheet) adjacent to or below the parameter. All times are UTC. Typically, times and time periods will be given in the format shown in the following examples:
Time:
06W
= 0600 UTC Wednesday;
21F = 2100 UTC Friday.
Time period:
12-18Th
= 1200 UTC to 1800 UTC Thursday;
18M-06T = 1800 UTC Monday to 0600 UTC Tuesday
For some parameters, a list of categories is provided. For these parameters, the sum of all of the categorical probabilities MUST add to 100%, no more no less. For the remaining parameters, any probability from 0 to 100 is acceptable.
Temperature:
Four temperature parameters (T1, T2, T3 and T4) are forecast
for each city. For each temperature parameter, there are eleven
temperature categories to which a probability (from 0 to
100) must be assigned. Each category spans a temperature interval of
three degrees. The temperature value shown in each
category on the spreadsheet is the center value of each three-degree
interval. The sum of the
probabilities in the eleven categories must add to 100%.
Precipitation:
Probability of precipitation (PoP1 and PoP2):
Probability of at least .01 inches in the specified time period.
Amount (Precip): There are
eight categories of precipitation amount (in hundredths of inches) for a given
time period to which a
probability (from 0 to 100) must be assigned. As with the
temperature categories, the sum of the categorical
probabilities must add
to 100%.
Duration (% TimePrecip): %
of the time that precipitation (not necessarily
measurable) is reported during the forecast time period.
Types
of Weather:
Dewpoint (Td): As
with temperature, there are eleven dewpoint
categories to which a probability (from 0 to 100) must be
assigned. Each category spans a temperature interval
of three degrees. The dewpoint value shown in
each category on the
spreadsheet is the center value of each three-degree
interval. The sum of the probabilities of each
category must
add to 100%.
Wind direction (Wind Dir): There are 13 categories of
wind direction. Each category spans 30°, except, of course, for the light
and
variable category. If the forecast is for a specific time,
the three-hour average (centered on the forecast time) is used for
verification (unless you are told otherwise). If the
forecast is for a time period, the average over the period is computed for
verification. Only values taken “on the
hour” are used in computing the averages. The
sum of the probabilities of each category must add to 100%.
Wind speed: Wind speed is divided into eleven categories.
If the forecast is for a specific time, the three-hour average (centered
on the forecast time) is used for verification (unless you
are told otherwise). If the forecast is for a time period, the average over
the period is computed for verification. The sum of the probabilities of each category must add to
100%.
Cloud cover (Avg cloud): If
the forecast is for a specific time, the three-hour average (centered on the
forecast time) is used for
verification (unless you are told otherwise). If the
forecast is for a time period, the average over the period is computed for
verification.
Individual reports are converted to tenths using the following values:
(OVC =>1.0, BKN => 0.7, SCT => 0.3, FEW => 0.1, CLR => 0.0)
Events:
Snow: any mention of snow in either the hourly or
special observations.
Thunderstorm: any mention of thunder in either the
hourly or special obs.
ZR/IP: any mention of any of FZRA/FZDZ/PL (freezing
rain/freezing drizzle/ice pellets) in hourly or special obs.
FOG: any report of fog at the station.
VIS: any hourly or special with visibility less than
or equal to the indicated threshold distance (in miles).
LTG: any remark of lightning (CC, IC, CG)
CIG: Cloud ceiling (base). Any hourly or
special with ceiling less than or equal to the indicated threshold altitude (in
feet).
Requires a BKN or OVC deck.
Additional
forecast parameters will be chosen as the forecast situation warrants
^ Introduction
Forecasts are entered onto an EXCEL97 spreadsheet and then sent via FTP to a clearing site where they are scored. If you are working with a UNIX/LINUX system, you may first have to install the program StarOffice onto your account in order for you to be able to work with the spreadsheet and enter your forecast. If you are working on a PC, then you should be able to open the spreadsheet without installing StarOffice.
On
each day that you forecast, the following steps must be taken:
1) Prior to the start of each class, an appropriate spreadsheet file for the
day’s forecast will be sent to your email account as an attachment. It will also be posted on the
instructor’s web site. Please
check the list of names and email addresses below to be sure that the address
listed for you is correct. The
spreadsheet file sent to you will be titled “date_name.xls” where
the actual date will replace the word “date”, e.g., the spreadsheet
for September 4 will be titled “0904_name.xls”
2) Open your email account and save the file. WATCH
WHERE YOU SAVE IT.
3) If you are working UNIX/LINUX, open StarOffice and
then open the spreadsheet file (m415forecast_sep4.xls) from within the StarOffice program.
4) Enter your forecast probabilities (in percent) on the spreadsheet.
5) Save the spreadsheet file as a “Text (tab
delimited)” file and insert the first six digits of your email name in
place of “name”. For
example, on September 4, someone who has the email address
“abc123@psu.edu” would save their file as
“0904abc123”. After
you have made the necessary changes in the name of the file and you complete
the “save as” command, the EXCEL program will ask you if you are
sure you want to do this. Click on
“yes” or “OK”. EXCEL will then ask you again - - -
click on “yes” or “OK” again. WATCH TO BE SURE YOU KNOW IN WHICH FOLDER YOU HAVE SAVED YOUR
FILE.
6) Now open the FTP program. Under the “general” tab, insert the following settings:
- For “Host Name/Address”, use “stokes.meteo.psu.edu”
- For User ID, use “m415”
- For the password, use “contest”
Then hit “return” or “OK” and go to the left side of the FTP window and find your forecast file that you saved on your local system. Select it and then click on the arrow that moves the file in the appropriate direction.
7)
If, after you have submitted your forecast, you realize that you need to update
your forecast, simply save the corrected file with “new” tacked on
the end of the file name. For example,
if the name of the original file was “0904abc123”, save the
corrected file as “0904abc123new” and then FTP the corrected file
to the clearing site in the same manner as described above.
8) If for some mysterious reason, you are unable to FTP your forecast, open
your e-mail account and send the file (as an attachment) to
m415mgr@mail.meteo.psu.edu.
|
Team post-mortems (for forecasts
made on Tuesdays) are presented on Thursdays |
Date of Tuesday forecast |
|
|
|
|
Team A: Jeff Kron Pete Mangione |
14 October
4 November |
|
|
|
|
Team B: Chris Beatty Ben Legg |
16 September 28 October |
|
|
|
|
Team C: Dayna Sherwood Kristina Baker |
30 September 18 November |
|
|
|
|
Team D: Pete Cannella Jonathan Pacheco |
23 September 11 November |
|
|
|
|
Team E: Josh Meister Tim Silfies
Ed Skirkie |
7 October
2 December |
|
|
|
|
Team F: Chris Nixon Jim Rourke Katie Will |
21 October
9 September |
|
|
|
|
|
|
|
|
|
1. "The only thing we learn from
history - - - is that we do not learn from history." Milton
Friedman
2. TRUE
FACTS....
Only in
America......can a pizza get to your house faster than an ambulance.
Only in
America......are there handicap parking places in front of a
skating rink.
Only in
America......do drugstores make the sick walk all the way to the
back of the store to get their prescriptions while healthy
people can buy
cigarettes at
the front.
Only in
America......do people order double cheeseburgers, large fries, and
a diet coke.
Only in
America......do banks leave both doors open and then chain the pens
to the counters.
Only in
America......do we leave cars worth thousands of dollars in the
driveway and put our useless junk in the garage.
Only in
America......do we use answering machines to screen calls and then
have call waiting so we won't miss a call from someone we
didn't want to
talk to in
the first place.
Only in
America......do we buy hot dogs in packages of ten and buns in
packages of eight.
Only in
America......do we use the word 'politics' to describe the process
so well: 'Poli' in Latin
meaning 'many' and 'tics' meaning 'bloodsucking
creatures'.
Only in
America......do they have drive-up ATM machines with Braille
lettering.
EVER WONDER ......
Why the sun lightens
our hair, but darkens our skin?
Why women
can't put on mascara with their mouth closed?
Why don't you
ever see the headline "Psychic Wins Lottery"?
Why is
"abbreviated" such a long word?
Why is it
that doctors call what they do "practice"?
Why is it
that to stop Microsoft Windows, you have to click on "Start"?
Why is lemon
juice made with artificial flavor, and dishwashing liquid made
with real
lemons?
Why is the
man who invests all your money called a broker?
Why is the
time of day with the slowest traffic called rush hour?
Why isn't
there mouse-flavored cat food?
When dog food
is new and improved tasting, who tests it?
Why didn't
Noah swat those two mosquitoes?
Why do they
sterilize the needle for lethal injections?
You know that
indestructible black box that is used on airplanes? Why
don't they
make the whole plane out of that stuff?!
Why don't
sheep shrink when it rains?
Why are they
called apartments when they are all stuck together?
If con is the
opposite of pro, is Congress the opposite of progress?
If flying is
so safe, why do they call the airport the terminal?
ENGLISH....
If you ever feel stupid, then just read on. If you've learned
to speak fluent English, you must be a genius!
This little treatise on the lovely language we share is only
for the brave. Peruse at your leisure, English lovers.
Reasons why the English language is so hard to learn:
1) The bandage was wound around the wound.
2) The farm was used to produce produce.
3) The dump was so full that it had to refuse more refuse.
4) We must polish the Polish furniture.
5) He could lead if he would get the lead out.
6) The soldier decided to desert his dessert in the desert.
7) Since there is no time like the present, he thought it was
time to present the present.
8) A bass was painted on the head of the bass drum.
9) When shot at, the dove dove into the bushes.
10) I did not object to the object.
11) The insurance was invalid for the invalid.
12) There was a row among the oarsmen about how to row.
13) They were too close to the door to close it.
14) The buck does funny things when the does are present.
15) A seamstress and a sewer fell down into a sewer line.
16) To help with planting, the farmer taught his sow to sow.
17) The wind was too strong to wind the sail
18) After a number of injections my jaw got number.
19) Upon seeing the tear in the painting I shed a tear.
20) I had to subject the subject to a series of tests.
21) How can I intimate this to my most intimate friend?
There is no egg in eggplant nor ham in hamburger; neither
apple nor pine in pineapple. English muffins weren't invented
in England or French fries in France (Surprise!). Sweetmeats
are candies while sweetbreads,which
aren't sweet, are meat.
Quicksand works slowly, boxing rings are square and a guinea
pig is neither from Guinea nor is it a pig. And why is it that writers
write but fingers don't fing, grocers don't groce and hammers don't ham?
If the plural of tooth is teeth, why isn't the plural of booth beeth?
One goose, 2 geese. So one moose, 2 meese? Doesn't it seem crazy that
you can make amends but not one amend. If you have a
bunch of odds
and ends and get rid of all but one of them, what do you call it? Is it an
odd, or an end?
If teachers taught, why didn't preachers praught? If
a vegetarian eats
vegetables, what does a humanitarian eat? In what language do people
recite at a play and play at a recital? Ship by truck and send cargo by
ship? Have noses that run and feet that smell?
How can a slim chance and a fat chance be the same, while a wise man
and a wise guy are opposites? You have to marvel at the unique lunacy
of a language in which your house can burn up as it burns down, in
which you fill in a form by filling it out, and in which, an alarm goes
off by going on.
English was invented by people, not computers, and it reflects the
creativity of the human race, which, of course, is not a race at all.
That is why, when the stars are out, they are visible, but when the
lights are out, they are invisible.
P.S. - Why doesn't "Buick" rhyme with "quick"?
klb311@psu.edu
cmb390@psu.edu
pac165@psu.edu
jak445@psu.edu
bhl112@psu.edu
kzl107@psu.edu
pcm134@psu.edu
jtm228@psu.edu
cen115@psu.edu
jmp386@psu.edu
jrr197@psu.edu
dms418@psu.edu
trs171@psu.edu
ejs225@psu.edu
klw231@psu.edu
klb311@psu.edu,cmb390@psu.edu,pac165@psu.edu,jak445@psu.edu
** GENERAL
**
Comprehensive list of weather servers
on the net
Comprehensive list of numerical
models on the web
State College NWS MMM model fcst
output
from 11 different models
surface,
upper-air, satellite and radar data
FSL Datasets__realtime_and_archived
HPC quantitative
precipitation forecasts
Storm Prediction Center severe weather
reports
Storm Prediction Center Map Archives
current
warning/advisory map
NWS warnings, watches, advisories and storm
reports
Federal
Meteorological Handbook #1
weather symbols
chart
wind chill table
map and links to
state climatologists offices
COMET case studies
global weather
guidance and forecast archives
