Vous êtes sur la page 1sur 7

Student-Forecaster Performance Exercise on Tropical Cyclone1

Labuyo2
Joseph Basconcilloa, Paulo Goocoa, Nikos Peñarandaa,
Thet Htar Su Hlainga, Rhonalyn Vergaraa

Abstract. On September 3, 2009, Tropical Cyclone (TC) Labuyo entered the Philippine
Area of Responsibility (PAR) affecting most of the areas in the eastern part of the country.
The Philippine Atmospheric, Geophysical, and Astronomical Services Administration
(PAGASA) said that on September 6, 2009, TC Labuyo would leave the PAR and would go
along the south coast of Japan. This paper presents the forecasting performance exercise
on TC Labuyo of graduate students from the University of the Philippines-Diliman. This
study does not provide an alternative forecast output of TC Labuyo but rather a mere
reiteration of synoptic methods and techniques for the purpose of student-forecasting
performance exercise. Employing Manual Forecasting Method used in Myanmar and
Persistence Method, this study proves that the performance skill of student-forecasters is
relatively high since the process of generating TC Labuyo forecast is independent from
other forecasts made by weather agencies maintaining consistency, reliability and had
developed critical thinking, motivation, and forecasting skill among the student-
forecasters.

Introduction. Tropical cyclones are the most Philippine Atmospheric, Geophysical, and
frequent and most devastating natural Astronomical Services Administration
disaster posed by our environment. 30% to (PAGASA) said that on September 5, 2009,
38% of the total number of tropical cyclones TC Labuyo would leave the PAR and would
hit the western North Pacific Region where go along the south coast of Japan. Within this
the Philippines belongs (Gonzales, 1994). It is short span of stay in PAR Labuyo had great
no doubt that PAGASA has its hands full impacts in the economy and livelihood of the
throughout the year preparing, studying, and people within the affected areas.
forecasting possible behaviour of TC visiting
the country and its accompanying torrential This paper presents the forecasting
rains and strong winds, and providing daily performance exercise of graduate students
weather forecasts for various purposes. from the University of the Philippines-Diliman.
Under the structure of World Meteorological
Agency, PAGASA is designated as the Figure 1. TC Labuyo left the PAR on
National Weather Service of the Philippines. September 6, 2009

On September 3, 2009, Tropical Cyclone 1


(TC) Labuyo2 entered the Philippine Area of
Responsibility (PAR) affecting most of the
areas in the eastern part of the country. The






























































a

Institute
of
Environmental
Science
and
Meteorology,

University
of
the
Philippines‐Diliman

1

Under
the
Saffir‐Simpson
Scale,
Labuyo
fell
under
the

Tropical
Storm
category.
However,
for
the
purpose
of

brevity,
Tropical
Cyclone
(TC)
is
used
in
this
paper.

2

Labuyo
has
an
international
name
Dujuan.

According to Market (2006), forecasting (1) The TC of interest is within the
exercises encourages critical thinking among agency’s jurisdiction or area of
forecasters and as such this study does not responsibility;
provide an alternative forecast other than the (2) The services and products of such
public bulleting of TC Labuyo from PAGASA agency and/or provider are readily
but rather a mere reiteration of synoptic accessible;
methods and techniques for the purpose of (3) The services and products by such
student-forecasting exercise. agencies are timely and current;
(4) The agency is technically capable and
Data Repositories. The data set includes is reputable.
analysis charts and model outputs pertaining
to mean sea level pressure, wind fields at Figure 2. Spatial Distributions of Designated
various levels, forecast tracks and storm logs.
Satellite images were also obtained for
validation purposes. These images only
provided a visual description of TC Labuyo.
The data set were obtained from both local
and foreign weather agencies and providers
below:

• Philippine Atmospheric, Geophysical,


and Astronomical Services
Administration (PAGASA);
• Japan Meteorological Agency (JMA) /
Regional Specialized Meteorological
Center (RSMC)-Tokyo;
• Bureau of Meteorology Australia Agencies (Source: Comet Program, 2009)
(BOM);
• Joint Typhoon Warning Center Table 1. Overview of Data Sources
(JTWC); Agency Brief Key Products Key NWP/ Models Key TC Forecast Method
• Typhoon2000.com (an independent mslp, streamline charts;
aviation and shipping
local weather blog site); and PAGASA National Weather Service (Philippines)
forecasts (local only); MM5, HRM, Wave, ETA, Persistence; model

• Naval Research Laboratory (NRL Satellite data; Tropical


Cyclone Track Archive
Storm Surge outputs (GFS, etc.)a

Tropics). (PAR)
Various Forecast Charts for
GASP, LASP, TXLAPS,
Australia and global domain
National Weather Service (Australia); WAM, Storm Surge,
BOM (ie. Sub-surface N/A
Figure 2 shows the map of agencies TCWC (South Pacific & SE Indian Ocean)
Temperature Analysis,
Coupled Ocean-
Atmosphere, OCF
designated by World Meteorological Agency SST)
Satellite Data (MTSAT), Dvorak, MicrowaveSat.
One-week Ensemble, HR
to oversee the tropical cyclone nomenclature JMA
Japan Meteorological Agency; RSMC mslp, TC Forecasts,
Global Model, GSM,
Imagery, Typhoon
(West Pacific Region) Volcanic Ash Advisories, Ensemble prediction
and to provide advisories and bulletins with Climate System Forecasts
Mesoscale Model
systems
TC Forecast Charts, TC
up-to-date first level basic meteorological US Government Agency; “Unofficial Formation Alerts, TC
1-minute average
JTWC CONW (ensemble) estimation of max wind;
information on all tropical cyclones to their TCWC” (Pacific and Indian Ocean Basins) Archive, Satellite data,
Prognostic Reasonings
Dvorak, ensemble models
respective regions. NRL Tropics US Government Agency Processed Satellite Data N/A N/A
An independent tropical cyclone well-
Poolsvarious data from
Typhoons2000.com maintained blog site based in Naga, City, N/A N/A
different agencies.
Table 1 provides information on various Philippines

meteorological data available for procurement


during tropical cyclone onset from different Project Workstation. The student-
weather agencies. forecasters used a system in processing
synoptic weather observations and other
The agencies and providers were selected dataset from other weather agencies. The
according to the following criteria: system was designed as organic. In analogy,
each task served as an organ to administer
the whole process of generating a final and are known worldwide and are used for
reliable forecast of TC Labuyo. operational and research purposes. All of the
models are retrofitted to simulate Philippine
weather conditions.

A brief discussion of each model is provided


below:

1. The MM5 is a limited-area, non-


hydrostatic or hydrostatic, terrain-
following sigma coordinate model
designed to simulate or predict
mesoscale and regional-scale
atmospheric circulation (MM5
Community Model, 2003).

The PAGASA forecast uses the


version V3 of the MM5. The PAGASA
MM5 model consists of two domains
which is the outer domain and the
nested domain. The former is used to
simulate large scale features that are
affecting the Philippine region and
Figure 3. Flowchart of Workstation Process
adjacent areas such as monsoons,
tropical cyclones, and other tropical
The “Library” provides archival and storage
weather disturbances. While the latter
functions to the workstation. Forecasters can
is designed to simulate small scale
easily retrieve records of past tropical
features (PAGASA, 2009).
cyclones and weather events. In the same
manner, observed weather data can also be 2. The model is based on equations
stored in the library for future references. The
describing the evolution of variables
storage includes analysis charts and model
such as temperature, wind speed,
outputs, synoptic observation, forecast tracks
humidity, and pressure (PAGASA,
and storm logs, and satellite images.
2009). The HRM is a hydrostatic and
uses a hybrid sigma coordinate
Research outputs and credible journals are
system.
also included in the system to develop
forecast performance among student-
The available outputs from the
forecasters. Two-methods were employed in
PAGASA using HRM are rainfall,
this study – Synoptic and Persistence temperature, winds, and pressure at
Methods. However, PAGASA uses different
the surface level. As for the
numerical weather predictions in supplement
atmosphere level, the outputs are
with synoptic and persistence methods.
temperature, moisture, winds, and
geopotential height. For timeliness,
PAGASA uses four numerical weather
PAGASA utilizes the Global Spectral
prediction models to forecast tropical
Model (GSM) with the HRM.
cyclones and simulate weather conditions in
the Philippines. These models are (1) the
3. The Eta model is designed to
Mesoscale Model Version 5 (MM5), (2) the
simulate weather conditions covering
High Resolution Model (HRM), (3) the Eta the entire Philippines and adjacent
Model, and (4) the Wave Model. The models
areas (101˚E to 149˚E and 6˚S to
28˚N) at a horizontal resolution of 28- Synoptic Analysis on Tropical Cyclone
km centered at 11˚N latitude and • Presence of a Pre-existing disturbance
125˚E longitude. It uses the output of • Large elliptical or circular wind
Japanese GSM for the initial circulation on ITCZ
atmospheric conditions of the model • Persistence of low-level easterlies to
(PAGASA, 2009). North of the system
• Low-level positive relative vorticity
4. The wave model is a • Low-level convergence and upper-
numerical/spectral model which level divergence. An upper-level
evolved from the time-dependent anticyclone above the system and a
energy balance equation of the wave Tropical Upper Tropospheric Trough
spectrum. Such models treat the (TUTT) to the west of the system are
various processes that control ocean indicative of an intensifying system.
waves. The wave model adapted by • Tropical Cyclones move in the
PAGASA was developed by direction and speed of the Steering
METEOFRANCE called VAG model. Current
It is a 2nd generation wave model
applicable for deep water used for Physical Analysis on Tropical Cyclone
operational purposes over the Atlantic • Large positive zonal wind shear
Ocean and Mediterranean Sea. poleward and negative zonal shear
Winds from the GSM are used to drive equatorward; southerly shear to the
the Philippine waters and obtain wind- west and northerly shear to the east
generated waves for 72 hours over the region
(PAGASA, 2009). • Low-level wind surges that propagate
inward towards the center of the
Synoptic Method. This performance exercise disturbance
used the Synoptic or Manual Forecasting • Unusual pressure fall in and around
method employed in the country of Myanmar. the center of deepening easterly wave
Myanmar is located in the Indian Ocean • Interaction with other low pressure
basin, which is also frequented by tropical systems
cyclones.
In addition to the guidelines, the use of the
Manual Forecasting Method employs three
faculties namely:

• Observational tools;
• Convectional observation; and
• Satellite observation.

In conducting the forecast, the following data


was used:
(1) Surface chart (isobaric analysis);
(2) Contour analysis (850 hPa to 200
hPa);
(3) Wind analysis (isogon, isotach and
Figure 4. Global Distribution of Observed streamline analysis); and
Tropical Cyclones Track (1851-2006) (4) Isollabaric chart analysis and other
(Source: Comet Program, 2009) observational tools.

Below are guidelines on how the The weather elements mentioned above was
method was conducted: analyzed by identifying weather systems
present during the forecast period. This was Philippines. This situation can be seen in
done by tracing the isobar lines, in the case of slides 13 to 15 (refer to manual fc.ppt).
surface pressure; streamlines, for the wind,
and etc., until all systems become apparent. Persistence Method. A simple equation is
As a result, the location, intensity and used in determining the possible tracks of TC
movement of tropical storm Labuyo became Labuyo. The Persistence Method equation is
possible to determine. A complete list of the given by:
analysis charts can be referred to by browsing
the PowerPoint document, manual fc.ppt, Φ24 = Φ12 – 2 (Φ12 – Φ0) (1)
located in the forecast folder. Other data can λ24 = λ12 – 2 (λ12 – λ0) (2)
be referred to in the Weather Data folder.
At 00 UTC on 3 September, Labuyo was Where
located about 18N and 129E (NE of Φ24 is the forecasted 24-hour latitudinal
Philippines). One possible forecast based on location of the tropical cyclone;
isobaric charts point out that Labuyo might Φ12 is the latitudinal location of the tropical
have moved towards the Philippines. The cyclone 12 hours before the observed
lowest mean sea level pressures extend from location;
the east of the Philippines, to the South China Φ0 is the observed latitudinal location;
Sea and, all the way to India. This particular λ24 is the forecasted 24-hour longitudinal
situation can be seen in slides 1 to 2 (refer to location of the tropical cyclone;
manual fc.ppt). Low-level convergence (850 λ12 is the longitudinal location of the tropical
hPa) and upper level divergence (200 hPa) cyclone 12 hours before the observed
was evident over Philippine Sea along the location;
trough line and ridge line. It can be seen in λ0 is the observed longitudinal location.
slides 3 to 4 (refer to manual fc.ppt).
From this equation, a forecast of possible
At 00 UTC on 4 September, Labuyo appears tracks was made for TC Labuyo.
to re-curve as a prominent trough moved
eastward. Slides 5 to 7 illustrated that the Φ24 Lat λ24 Lon
streamline pattern gave a distinct impression (N) (E)
that Labuyo was about to cross the axis of the 03/1800Z 19.3 130.1
subtropical ridge. (Refer to manual fc.ppt). 04/0000Z 18.0 127.7
04/0600Z 16.0 123.4
At 00 UTC on 5 September, Labuyo
continued moving to an eastward direction. At 04/1200Z 15.6 123.5
this point, it appeared that Labuyo was at its 04/1800Z 16.8 128.0
easternmost position as shown in slide 8 to 9 05/0000Z 17.5 130.5
(refer to manual fc.ppt). One forecast showed
05/0600Z 21.1 133.1
a possibility for Labuyo to take an east-
northeastward or even eastward track under 05/1200Z 21.8 134.5
the steering influence of a northeasterly flow 05/1800Z 20.0 134.8
that is southeast of the anticyclone centered 06/0000Z 22.4 136.5
over the East of China. This condition is Table 2. Possible Tracks of TC Labuyo using
shown in slides 9 to 12 (refer to manual Persistence Method
fc.ppt).
Although the Persistence Method equation is
At 00 UTC on 6 September, Labuyo have very crude in estimating the possible tracks of
moved northeastward as the anticyclone a tropical cyclone, it is nevertheless helpful in
originally over the East of China had providing a better picture of the behavior of
disappeared. Labuyo appeared to be exiting the tropical cyclone with respect to nearby
the PAR and posed no direct threat to the
weather systems such as subtropical high Contemplating between numerical weather
pressure areas and ridges. It is interesting to prediction and persistence method,
note that such weather systems are also Mittermaier (2008) said that,
apparent in the synoptic method.
“So what is the appropriate choice of
reference? To judge the value of NWP
forecasting systems, we should ask the
following question: Is it possible to
demonstrate that a computationally
expensive NWP model provides
significant skill and added value over a
simple persistence forecast (pp.1030)...”

There is no compartmentalization or mutual


exclusivity between different methods in
forecasting. Performance skill marks this
difference relating it to the ability of the
forecaster and the reliability of its forecasts.
Figure 5. Plotted Forecast Track of TC Conclusion. Although another forecast was
Labuyo based on 1800-0600 UTC intervals generated from persistence method, this
using Persistence Method study does not intend to contest previous
forecast made by various weather agencies.
This study simply reiterates the ability of
student-forecasters to produce reliable TC
forecasts with the aid of the designed project
workstation and available meteorological
dataset.

The human element is crucial in weather


forecasting (Doswell and Lemon, 1981;
Doswell, 1986). Tropical cyclone forecasters
should be well motivated, educated, trained
and should have a talent for tropical cyclone
weather and weather forecasting. Forecasters
should have developed critical thinking to
Figure 6. Plotted Forecast Track of TC assess the variability of dataset at hand. In
Labuyo based on 0000-1200 UTC intervals addition, forecasters should be capable of
using Persistence Method explaining and defining the reasons for their
own forecasts or those of others to an end
Performance Skill. According to Mittermaier user who may or may not have any
(2008), (forecasting) skill is defined as the meteorological background.
ability to produce actual forecast performance
relative to the reference forecast In conclusion, this study proves that the
performance. A degree of disconnection from performance skill of the student-forecasters is
other forecasts must be maintained to keep relatively high since the process of generating
the reliability and consistency of the forecast TC Labuyo forecast is independent from other
performance. forecasts made by weather agencies
maintaining consistency, reliability and had
developed critical thinking, motivation, and
forecasting skill among the student- • www.pagasa.dost.gov.ph/wb/wb.html
forecasters. • www.wmo.int/pages/prog/www/tcp/index_en.html

Acknowledgement: The authors would like to


give thanks to Dr. Leoncio Amadore for
valuable support.

References:
_____, (2003). An overview of MM5 modeling system.
MM5 Community Model. Accessed on September 21,
2009, from http://www.mmm.ucar.edu/mm5/overview.

_____,(2009). Model description and configuration.


Philippine Atmospheric Geophysical and Astronomical
Services Administration. Accessed on September 21,
2009, from http://www.pagasa.dost.gov.ph.

______, (2009). Chapter 10: Tropical Cyclones. COMET


Program. Accessed on September 15, 2009 from
http://www.meted.ucar.edu/tropical/.

Doswell, C. A., (1986). The human element in weather


forecasting. National. Weather. Digest, 11, 6-18.

Doswell, C. A., and Lemon, L.R., (1981). Forecaster


training - A review and analysis. Bull. Amer. Meteor.
Soc., 62, 983-988.

Gonzales, E. (1994). Tropical Cyclones and Storm


Surges. Paper presented at the National Conference on
Natural Disaster Mitigation, Oct 19-21, 2009. Quezon
City: Philippines

Lam, C. Y., (1991). A Brief synoptic Discussion of


Spectrum Tropical Cyclones. Submitted for publication
in WMO Bulletin, Vol. 40, No. 1.

Market, S. (2006). The Impact of Writing Area Forecast


Discussions on Student Forecaster Performance.
Weather and Forecasting. Volume 21, pp. 104-108.

Mittermaier, M. (2008). The Potential Impact of Using


Persistence as a Reference Forecast on Perceive
Forecast Skill. Weather and Forecasting. Volume 23,
pp. 1022-1031.

Tun, L., (1986). Forecaster Handbook for Myanmar.


National Meteorology Center (NMC), DMH:Yangon,
Myanmar.

WWW Links
• www.bom.gov.au/weather/cyclone/about/tropical-
cyclone-intensity.shtml.
• www.nrlmry.navy.mil/
• www.bom.gov.au/nmoc/ab_nmc_op.shtml
• metocph.nmci.navy.mil/jtwc/menu/JTFAQ.html
• www.jma.go.jp/jma/jma-eng/jma-center/rsmc-hp-
pub-eg/techrev.htm

Vous aimerez peut-être aussi