Weather Forecasting Models Methods and Applications IJERTV2IS120198

International Journal of Engineering Research & Technology (IJERT)
ISSN: 2278-0181
Vol. 2 Issue 12, December - 2013
Weather Forecasting Models, Methods and Applications

AA
Iseh. A. J. * Woma. T. Y.1,2

1
1. Department of Pure & Applied Physics, Federal University Wukari, Taraba State.
P. M. B. 1020 Wukari, Taraba State.
2. Department of Pure & Applied Physics, Federal University Wukari, Taraba State.
P. M. B. 1020 Wukari, Taraba State.
ABSTRACT weather data, forecasting methods and

Weather forecasting modelling is a applications.
computer program that provides
meteorological information for future 1.0 INTRODUCTION
times at given locations. In modern
forecasting models, Numerical Weather
Prediction is mostly applied and this Modern society’s ever-increasing demand
means, “a set of simplified equations used for more accurate weather forecasts is
to calculate changes in atmospheric evident to most people. The spectrum of
conditions”. The act of writing these needs for weather predictions ranges from
equations, imposing the boundary the general public’s desire to know if for
conditions and solving them using super instance, the weekend will permit an
RT
computers, is known as numerical outing at the beach, or an organization’s
modelling. An example of such equations rally, or an outdoor wedding reception.
is the Hypsometric equation given as PI = Such diverse industries as airlines and fruit
IJE
POexp–gz/RT. Computerized numerical growers depend heavily on accurate

models are designed for different intervals weather forecasts to have an idea of what
which are known as global models under their next schedule of flight would appear
which we have long range forecast and to be or if the weather will be suitable for
medium range forecast, and regional harvesting. In addition, in developed
models under which we have the short countries, the designs of buildings, and
range forecast. The methods include many industrial facilities rely heavily on a
persistence, climatologic, looking at the sound knowledge of the atmosphere.
sky, use of barometer, nowcasting, use of Weather forecasting can be defined as the
forecasting models, analogue and act of predicting future weather conditions
ensemble forecasting. Forecasting could be or an attempt to indicate the weather
applied in air traffic, severe weather alerts, conditions which are likely to occur.
marine, agriculture, utility companies,
private sector and military application. Weather forecasting is the application of
Weather forecasting is a complex and Science and Technology to predict the
challenging science that depends on the state of the atmosphere for a future time
efficient interplay of weather observation, and a given location. Human beings have
data analysis by meteorologist and attempted to predict the weather
computers, and rapid communication informally for millennia, and formally
system. since at least the nineteenth century.
Weather forecasts are made by collecting
Key words: Weather, weather qualitative data about the current state of
prediction, forecast, forecasting models, the atmosphere and using scientific
understanding of atmospheric processes to
IJERTV2IS120198 www.ijert.org 1945

ISSN: 2278-0181
project how the atmosphere will evolve 2.0 HOW MODELS CREATE
within the next few hours.
FORECASTS
Once, an all-human endeavour based
mainly upon changes in barometric 2.1 Data collection
pressure, current weather conditions and Since invention of the first weather
sky conditions, forecast models are now instruments in the seventeenth century
used to determine future conditions. A weather observation has undergone
model, in this context, is a computer considerable refinement. Denser
program that produces meteorological monitoring networks, more sophisticated
information for future times at given instruments and communication systems,
positions and altitudes. The horizontal and better-trained weather observers, have
domain of a model is either global, produced an increasingly detailed, reliable
covering the entire earth, or regional, and representative record of weather and
covering only part of the earth. Regional climate. In weather forecasting, data
models also are known as limited area collection has been divided into two
models. Human input is still required to categories namely:
pick the best possible forecast model to
i. Surface weather observations
base the forecast upon, which involves
ii. Upper-air weather observations.
pattern recognition skills, knowledge of
model performance and knowledge of
model biases. The chaotic nature of the Surface Weather Observations
atmosphere, error involved in measuring Surface weather observations are the
the initial conditions, an incomplete
RT
fundamental data used for safety as well as
understanding of atmospheric processes climatological reasons to forecast weather
mean that forecast become less accurate as and issue warnings worldwide. They can
the difference in current time and the time
IJE
be taken manually by a weather observer,

for which the forecast is being made by computer through observers to augment
increases. the otherwise automated weather station.
There are a variety of end users to weather Referring to Lutgens and TarBuck (1989),
forecasts. Weather warnings are important a vast network of weather stations required
forecasts because they are used to protect to produce a weather chart will encompass
life and property. Forecasts based on enough to be useful for short-range
temperature and precipitation are forecasts. On a global scale, the World
important to agriculture, and therefore to meteorological organization, which
commodity traders within stock markets. consists of over 130 nations, is responsible
Temperature forecasts are also used by for gathering the needed data and
utility companies to estimate demand over producing some general prognostic charts.
coming days. On an everyday basis people
use weather forecasts to determine what to Surface weather observations of
wear on a given day. Since in recent time atmospheric pressure, temperature, wind
in Uyo – Nigeria for example, outdoor speed and direction, humidity,
activities are severely curtailed by heavy precipitations are made near the earth’s
rains, forecasts can be used to plan surface by trained observers or automatic
activities around these events, and to plan weather stations. The World
ahead and survive them. meteorological Organization acts to
standardize the instrumentation, observing
practices and timing of those observations
worldwide. By international agreement,
the regular synoptic observations are made
every six hours beginning at midnight

ISSN: 2278-0181
Greenwish Mean Time (0000GMT, of points where there are processed by a

0600GMT, 1200GMT and 1800GMT) Central weather analysis organization.
each day. In addition, observations for Three locations have been designated by
aviation purposes are made at many the World Meteorological Organization as
airports every hour, or more often if the World Meteorological Centers, these are
weather is changing rapidly. Specialized located in Melbourne, Australia; Moscow,
observations may also be made under Russia; and Washington, D.C. U.S.A. In
certain conditions for agricultural, addition, most countries maintain national
industrial, research, or other purposes centers, where the basic weather needs of
(Miller and Thompson, 1975). the domestic economy are met. In Nigeria,
the National Meteorological Center is
located in Abuja.
Upper-air weather observations
Measurements of temperature, humidity 2.1.1 Data assimilation and
and wind above the surface are found by analysis
launching radiosondes on weather In order to do their work, most numerical
balloons. Radiosonde is a unit for use in models look at the atmosphere as a series
weather balloons that measures various of boxes. In the middle of each box is a
atmospheric parameters and transits them point for which the model actually
to a fixed receiver. Radiosondes may calculates weather variables and makes
operate at a radio frequency of 403MHz or forecasts. The result of this three
1680MHz and both types may be adjusted dimensional boxing up of the atmosphere
slightly higher or lower as required. is known as the grid; the point in the
RT
Mohan and Morgan (1991) states that, the middle is the grid point, and the distance
instrument transmits to the ground station between one point and another is called the
vertical profiles of air temperature, grid spacing (Ackerman and Knox, 2003).
IJE
pressure, and relative humidity up to an

altitude of about 30km. In addition, winds Grid point models of the atmosphere can
at various levels are computed by tracking get fussy when the data in the initial
the balloons with a radio direction finding conditions is not obtained at exactly the
antenna. location of the grid point. Also, the process
of creating an evenly spaced data set from
Upper-air weather data are also obtained irregularly spaced observations is called
by aircraft, dropwind sondes, radar, and interpolation.
satellites. Increasingly, data from weather
satellites are being used because of their Ackerman and Knox (2003) then say that,
almost global coverage. Although their the multiple jobs of interpolating and
visible light images are very useful for smoothing the data for use in numerical
forecasters to see development of clouds, models are collectively called data
little of this information can be used by assimilation.
numerical weather prediction models. The During the data assimilation process,
infrared data however, can be used as it information gained from the observations
gives information on the temperature at
is used in conjunction with a numerical
the surface and cloud tops. Individual model’s most recent forecast for the time
clouds can also be tracked from one time that observations were made, since this
to the next to provide information on wind contains information from previous
direction and strength at the clouds
observations. This is used to produce a
steering level. three-dimensional representation of the
Miller and Thompson (1975) agrees that, temperature, moisture and wind called a
the observations are collected at a number meteorological analysis. This is the models
estimate of the current state of the

ISSN: 2278-0181
atmosphere. Data assimilation proceeds by Referring to the work of Houghton (1986),

analysis cycles. In each analysis cycle, the task of writing the equations and the
observations of the current (and possibly, boundary conditions in a suitable form and
past) state of a system are combined with then of solving them with high speed
the result from and mathematical model digital computers is known as numerical
(the forecast) to produce an analysis, modeling. By comparing the behaviour of
which is considered as “the best” estimate the model with that of the real atmosphere,
of the current state of the system. This is the validity of the procedures employed by
called the analysis step. Essentially, the the model is tested. The most important
analysis step tries to balance the application of numerical modeling is the
uncertainty in the data and in the forecast. development of methods sufficiently
The model is then advanced in time and its reliable and sufficiently fast to be used in
result becomes the forecast in the next routine weather forecasting.
analysis cycle.
There are numerous equations employed in
2.1.2 Numerical weather prediction this work of forecasting models one of
which is the hypsometric equation which
Linacre and Geerts (1997) define can be derived from the hydrostatic
Numerical Weather prediction (NWP) as a equation written as
simplified set of equations called the
primitive equation used to calculate g
dP = -P dZ - (1)
changes of conditions. Modern weather RT
forecasting relies heavily on numerical We can simplify equation (1) by dividing
weather prediction.
RT
through with P to have
According to Lutgens and TarBuck (1989),
dP −g
the word “numerical” is misleading, for all = dZ - (2)
P RT
IJE
types of weather forecasting are based on

some quantitative data and therefore could Integrating the right hand side of equation
fit under this heading. Numerical weather (2) from P0 to P1 and the left hand side
prediction is based on the fact that the from Z0 to Z1, we have:
gases of the atmosphere obey a number of
known physical principles. Ideally, these P 1 dP −g Z1
physical laws can be used to predict the P0 P
= Z
dZ + C
RT 0
future state of the atmosphere, given the
current conditions. This situation is Let C = 0
analogues to predicting future position of P1 −g Z1
the moon based on physical laws and the InP P0 = Z Z0
RT
knowledge of its current position. Still, the
−g
large number of variables that must be → InP1 – InP0 = (Z1 – Z0)
included when considering the dynamic RT
atmosphere makes this task extremely −g
difficult. → In (P1/P0) = (Z1 – Z0) - (3)
RT
Manipulating the huge data sets and Taking the exponent of both sides of
performing the complex calculations equation (3), we have:
necessary to do this (weather prediction)
on resolution fine enough to make the −g
𝑒 In (P 1 P0)
= 𝑒[ RT (Z 1 – Z 0 )]
result useful requires the use of some of
the most powerful supercomputers. −g
P1 P0 = 𝑒[ RT (Z 1 – Z 0 )]

ISSN: 2278-0181
−g which are too slow to be important in short

→ P1 = P0𝑒 [ RT (Z 1 – Z 0 )]
range weather forecasting.
At sea level, Z0 = 0. Thus, Meteorological bureaux now regularly

provide seasonal outlook, with an accuracy
–g Z 1 notably enhanced by increased
P1 = P0𝑒 [ RT ] - (4) understanding of the relevance of the
southern oscillation, indicated by the sea
Equation (4) is the Hypsometric equation,
surface temperatures, the strength of the
which gives the pressure PI of the
trade winds, the location of areas of day
atmosphere at a particular height Z1,
convention across the tropical Pacific
Where P0 is the pressure at the sea level, Ocean, and the depth of themocline.
g is acceleration due to gravity. But of all these three types of numerical

weather prediction, Mohan and Morgan
R is the molar gas constant and (1991) say that actually, when viewed with
the objectivity of statistical analysis, short-
T is the temperature at the surface range weather forecasting is surprisingly
more accurate probably because of the
Mohan and Morgan (1991) agree with
short range of time within which this
Barry and Chorley (1992) that, some
forecast is made for.
computerized numerical models of the
atmosphere are designed to operate over The most modern models of weather data
different spatial scales depending on the processing systems for the two categories
RT
forecast range. For medium range forecasts of numerical models are as follows:
(up to 10 days), observational data are fed
into the computer from all over the globe, Global models: Some of the better-known
IJE
since within that forecast range a weather global numerical models are:
system may travel long distances. On the
other hand, for short-range forecasts (up to 1. Global Forecast System (GFS) –
3 days), the model utilizes data drawn Developed by the National
from a more restricted region of the globe. Organization for the Atmosphere in
Compared to a global model, a regional America. Output is freely available.
model offers the advantage of greater 2. NOGAPS – Developed by the US
resolution of data over a smaller area of Navy to compare with the GFS
interest.
3. Global Environmental Multi-scale
As it is, there are basically three types of Model (GEM) – Developed by the
numerical weather prediction models meteorological service of Canada
namely: short-range forecasts, medium
range forecast (as briefly discussed by 4. European Centre for Medium
Mohan and Morgan, 1991; and Barry and Range Weather Forecasts
Chorley, 1992 above), and long range (ECMWF) – a model run by the
forecast. Europeans with limited availability
Thus, Linacre and Geerts(1997) describes 5. UKMO – Developed by the United
long range forecasts as descriptive Kingdom Meteorological Office.
forecasts which are made for times of ten Limited availability, but is hand
to thirty days or one to four months (a corrected by professional
seasonal outlook). Also, use is made of a forecasters.
dynamical numerical weather prediction
model which allows for oceanic processes, 6. GME – developed by the German
Weather Service

ISSN: 2278-0181
7. ARPEGE – developed by the 6. High Resolution Limited Area

French Weather Service, Meteo Model (HIRLAM)
France. 7. GEM – LAM – Global
Environmental Multi-scale Limited
8. Intermediate General Circulation Area Model
model (IGCM) – developed by 8. Aladin: The high resolution limited
members of the Department of area hydrostatic and non-
Meteorology at the University of hydrostatic model developed
Reading. operated by several European and
North African countries under the
Regional models: Some of the better-
leadership of Meteo-France.
known regional numerical models are:
9. COSMO: The COSMO Model,
1. The Weather Research and formerly known as LM, aLMD or
Forecasting (WRF) Model was LAMI, is a limited area non-
developed co-operatively by NCEP hydrostatic model developed
and the meteorological research within the framework of the
community. WRF has several consortium for small scale
configurations including: modelling (Germany, Switzerland,
Italy, Poland and Greece).
a. WRF – NMM: The ERF www.wikipedia.com.
Non-hydrostatic Mesoscale According to Linacre and Geerts (1997),
Model is the primary short- the advantage of numerical weather
term weather forecast prediction is that it avoids errors of human
RT
model for the United State, judgment in deriving the prognosis, and
can be steadily improved by enlarging the
b. AR-WRF: Advanced amount and reliability of input data, by
Research WRF developed
IJE
new understanding of the Physics of

primarily at the United State atmospheric change, and by faster, larger
National Center for computers.
Atmospheric Research
(NCAR)
2.1.3 Model Output Post Processing
2. The North American Mesocale
Model (NAM) The raw output is often modified before
3. Colorado State University for being presented as the forecast. This can
numerical simulations of be in the form of statistical techniques to
atmospheric meteorology and other remove known biases (a term used to
environmental phenomena on describe a tendency or preference towards
scales from metres to hundreds of a particular perspective, ideology or result)
kilometres in the model, or its adjustment to take into
4. MMS – The fifth Generation account consensus among other numerical
mesoscale model weather forecast. MOS or Model Output
5. The Advanced Region Prediction Statistics is a technique used to interpret
System (ARPS) – developed at the numerical model output and produce site-
University of Oklahoma. It is a specific guidance. This guidance is
comprehensive multi-scale non- presented in coded numerical form, and
hydrostatic simulation and can be obtained for nearly all National
prediction system that can be used Weather Service reporting stations.
for regional scale weather
prediction up to the tornadscale
simulation and prediction.

ISSN: 2278-0181
3.0 FORECASTING PROBLEMS i. Imperfect data: The data of

today’s numerical models still
An important goal of all scientific includes a large helping of
endeavour is to make accurate predictions. radiosonde observations. However,
The physicist or chemist who conducts an the number of radiosonde sites in
experiment in the laboratory does so in the the World over has actually
hope of discovering certain fundamental declined over the past few decades.
principles that can be used to predict the Developed countries in the world
outcome of other experiments based on today, spend more money in
those principles. In fact, most of the laws launching weather satellites than
of science are merely very accurate for boring weather balloons.
predictions concerning the outcome of Satellite data are global in average,
certain kinds of experiments. But few but researchers in data assimilation
physical scientists are faced with more are still trying to figure out how
complex or challenging prediction this data can be “digested” properly
problems than the meteorologist. by the models. In addition,
In the first place, the meteorological important meteorological features
laboratory covers the entire globe, so that still evade detection, especially
even the problem of measuring the present over the oceans. The model results
state of the atmosphere is tremendous. are only as good as the data in its
Furthermore, the surface of the earth is an initial conditions.
irregular combination of land and water, ii. Faulty “vision” and “fudges”:
each responding in a different way to the Today’s forecasts also involve an
RT
energy source – the sun. Then, too, the inevitable trade-off between
atmosphere itself is a mixture of gaseous, horizontal resolution and the length
liquid, and solid constituents, many of of the forecast. This is because fine
IJE
which affect the energy balance of the resolution means lots of point at
earth, one of them, water, is continually which to make calculations. This
changing its state. Also, the circulations of requires a lot of computer time. A
the atmosphere range in size from forecast well into the future also
extremely large ones, which may persist requires millions or billions more
for weeks or months, to minute whirls, calculations. If fine resolution is
with life spans of only a few seconds. combined with a long range
forecast, the task would choke the
According to Miller and Thompson (1975) fastest supercomputers today. One
and Ayado and Burt (2001) the problem of would not get forecasts for weeks.
forecasting then, involves an attempt to Future improvement in computing
observe, analyze and predict the many will help speed things up.
interrelationships between the solar energy In the meantime, however, some
source, the physical feature of the earth, models are still not able to pick or
and the properties and motions of the “see” small-scale phenomena such
atmosphere. This is the basis on which a clouds, raindrops, and
weather forecasts still go wrong today. snowflakes. To compensate for this
fuzzy “vision” of models, the
Ackerman and Knox (2003) points out computer code includes crude
reasons why forecasts still go wrong today approximations of what is not
by stating that the limitations which being seen. These are called
directly relates to today’s numerical parameterizations. Even though
forecast models are as follows: much science goes into them, these
approximations are nowhere close
to capturing the complicated reality

ISSN: 2278-0181
of the phenomena. This is because; when it is steady state, such as during the
the smallest scale phenomena are summer season in the tropics. This method
often the most daunting to of forecasting strongly depends upon the
understand. Therefore, it is not an presence of a stagnant weather pattern. It
insult to meteorologists’ abilities to can be useful in both short-range forecasts
say that parameterizations are and long-range forecasts.
“fudges” of the actual phenomena.
Persistence forecasts are used by local
iii. Chaos: It will be surprising to note forecasters in determining such events as
that, even if a supercomputer which the time of the arrival of a thunderstorm
could do quadrillions of that is moving toward their region.
calculations each second were to be Persistence forecasts do not account for
invented, no better forecasting changes that might occur in the intensity or
result would still be gotten. Brute in the path of a weather system, and they
force numerical weather do not predict the formation. Because of
forecasting with extremely fine these limitations and the rapidity with
resolution has its limits. which weather system change in most
geographical regions, persistence forecasts
The reason for these limits is a break down after twelve hours, or a day at
curious property of complex, most.
evolving systems like the
atmosphere. It is called “Sensitive 4.2 Climatology forecasting
dependence on initial conditions”, Whereas persistence forecasting is most
and is a hallmark of what is
RT
accurate over short periods (before factors
popularly known as chaos theory.
for change have had time to operate), the
Chaos in the atmosphere does not
best estimate of the weather a long time
mean that everything is a mess;
IJE
ahead is the average value of past

instead, it means that the
measurements there at that time of day and
atmosphere both in real life and in year.
a computer model may read very
differently to initial conditions that Climatology forecast relies on the
are only slightly different. observation that weather for a particular
day at a location does not change much
Because we do not know the from one year to the next. As a result, a
atmospheric conditions perfectly at long term average of weather on a certain
any time, chaos means that the day or month should be a good guess as
resemblance between a model’s the weather for that day or month. The
forecast and reality will be less and most obvious climatology forecast in this
less with each passing day. part of the world (Nigeria) is, “Cold in
December, warm in July (the popular July
4.0 METHODS OF WEATHER break)”. One does not need to be a
FORECASTING meteorologist to make that forecast.
Of course for weather forecast to exist Today’s numerical forecast methods still
there must be methods on which it is done. use climatological statistics as a “reality
These methods are as follows: check”. There make sure that the computer
4.1 Persistence forecasting models are not going off the deep end,
climatologically speaking.
Persistence forecasting is the easiest
method of forecasting which assumes a
continuation of the present. It relies upon
today’s conditions to forecast the weather

ISSN: 2278-0181
4.3 Looking at the sky necessitating site-specific forecasts.

Along with pressure tendency, use of the Included in this category are
sky condition is one of more important thunderstorms, gust fronts, tornadoes, high
winds especially along coasts, over lakes
weather parameters that can be used to
forecast weather in mountainous areas. and mountains, heavy snow and freezing
Thickening of cloud cover or the invasion precipitation. The development of radar
networks, new instruments and high speed
of a higher cloud deck is indicative of rain
in the near future. Morning fog portends communication links has provided a means
fair conditions, as rainy conditions are of issuing warnings of such phenomena.
preceded by wind or clouds, which prevent Several countries including Nigeria have
fog formation. The approach of a line of recently developed integrated satellite and
thunderstorm could indicate the approach radar systems to provide information on
of a cold front. Cloud free skies are the horizontal and vertical extent of
indicative of fair weather for the near thunderstorms, for example. Such data are,
future. The use of sky cover in weather supplemented by networks of automatic
prediction has led to various weather lore weather stations that measure wind,
over the centuries. temperature and humidity.
4.4 Use of a barometer
Nowcasting methods use highly automated
Using barometric pressure and the pressure computers and image analysis systems to
tendency (xthe change of pressure over integrate data from a variety of sources
time) has been used in forecasting since rapidly. Interpretation of the data displays
the late 19th century. The larger the change requires skilled personnel and or extensive
RT
in pressure, especially, if more than software to provide appropriate
2.54mmHg, the larger the change in information. The prompt forecasting of
weather can be expected. If the pressure wind shear and downburst hazards at
IJE
drop is rapid, a low-pressure system is airports is one example of the importance

approaching, and there is a greater chance of nowcasting procedures.
of rain. Rapid pressure rises are associated
with improving weather conditions, such 4.6 Use of Forecasting Models
as clearing skies. In the past, the human forecasters were
4.5 Nowcasting responsible for generating the entire
weather forecast based upon available
The forecasting of the weather within the
observation. Today, human input is
next six hours is often referred to as generally confined to choosing a model
nowcasting. In this time range, it is
based on various parameters, such as
possible to forecast smaller features such model biases and performance. Using a
as individual showers and thunderstorms consensus of forecast models, as well as
with reasonable accuracy, as well as other
ensemble members of the various models,
features too small to be resolved by a
can help reduce forecast error. However,
computer model. A human given the latest regardless how small the average error
radar, satellite and observational data will
becomes with any individual system, large
be able to make a better analysis of the
errors within any particular piece of
small scale features present and so will be guidance are still possible on any given
able to make a more accurate forecast for model run. Humans can use knowledge of
the following few hours. local effects, which may be too small in
Severe weather is typically short-lived size to be resolved by the model to add
(less than two hours) and, due to its information to the forecast.
mesoscale character (less than one hundred
kilometers), it affects local/regional areas

ISSN: 2278-0181
4.7 Analogue Forecasting The main problem with this method may
The analogue method is a complex way of well be the lack of complete enough
making a forecast, requiring the forecaster information, which also of course, limits
the usefulness of numerical weather
to remember a previous weather event
which is expected to be mimicked by an predictions.
upcoming event. The analogue forecaster’s 4.8 Ensemble Forecasting
task is to locate the date in history when
the weather is a perfect match, or Although a forecast model will predict
analogue, to today’s weather. Then the weather features evolving realistically into
forecast for tomorrow is simple – whatever the distant future, the errors in a forecast
happened in the day after the analogue will will inevitably grow with time due to the
be the weather for tomorrow. The forecast chaotic nature of the atmosphere and the
for the day after tomorrow is whatever inexactness of the initial observations. The
happened in the second day after the detail that can be given in a forecast
analogue, and so forth. therefore decreases with time as these
What makes it a difficult method to use is errors increase. These become a point
that, there is rarely a perfect analogue for when the errors are so large that the
an event in the future. In fact, no two forecast has no correlation with the actual
patterns or sequences of weather are ever state of the atmosphere.
identical. There may, for example, be five However, looking at a single forecast gives
reasonable analogues for a particular no indication of how likely that forecast is
month, but examination of the succeeding to be correct. Ensemble forecasting entails
RT
weather sequences might show mild, rainy the production of many forecasts in order
weather in two cases and cold spells in the to reflect the uncertainty into the initial
other three. In the preparation of the state of the atmosphere (due to the errors
IJE
forecast therefore, many factors, which can in the observations and insufficient
affect the weather trends, such as sea sampling). The uncertainty in the forecast
temperature and the extent or amount of can then be assessed by the range of
rainfall, have to be taken into different forecasts produced. However, the
consideration. simple logic behind ensemble forecasting
is that two runs of a model are not enough
Some call this type of forecasting pattern to base a forecast upon.
recognition. It remains a useful method of
observing rainfall over data voids such as Ensemble forecasts are increasingly being
oceans, as well as the forecasting of used for operational weather forecasting.
precipitation amounts and distribution in Ensemble forecasting requires a
the future. A similar method is used in sophisticated understanding of the
medium range forecasting, which is known atmosphere and computer models.
as teleconnections – when systems in other
locations are used to help pin down the 5.0 APPLICATIONS OF
location of another system within the WEATHER FORECAST
surrounding regime. Teleconnections are
The importance of accurate weather
used by forecasters today to make general
forecasts cannot be over emphasized as the
forecasts months into the future. While in
needs for them are always craved for in
short term forecasting, the pattern
virtually every aspect of life. These
recognition is still used by weather
forecasts can be applied in the following
forecasters to supplement today’s
areas:
computerized methods. But in the end the
complexities of weather, like human
personalities, defy simple categorization.

ISSN: 2278-0181
5.1 Severe weather alerts and required for takeoff, and to eliminate
advisories potential crosswinds.
A major part of modern weather 5.3 Marine
forecasting is the severe weather alerts and Commercial and recreational use of
advisories, which the national weather waterways can be limited significantly by
services issue in the case that severe or
wind direction and speed, wave periodicity
hazardous weather is expected. This is
and heights, tides, and precipitation. These
done to protect life and property. Some of factors can each influence the safety of
the most commonly known severe weather
marine transit. Consequently, a variety of
advisories are the severe thunderstorm and
codes have been established to efficiently
tornado warnings, as well as the recent transmit detailed marine weather forecasts
warnings about areas that are prone to
to vessel pilots through radio, for example
flood in some part of Nigeria by the
the MAFOR (Marine forecast).
National Meteorological Agency. Other
forms of these advisories include winter 5.4 Agriculture
weather, high wind, flood, tropical
cyclone, and fog. Severe weather Farmers rely on weather forecasts to
advisories and alerts are broadcast through decide what work to do on any particular
the media, including radio, using day. For example, drying hay is only
emergency systems as the Emergency feasible in dry weather. Prolonged periods
Alert System, which break into regular of dryness can ruin cotton, wheat, and corn
programming. crops. While crops can be ruined by
drought, their dried remains can be used as
RT
5.2 Air Traffic a cattle feed substitute in the form of
silage. Frosts and freezes play havoc with
Because the aviation industry is especially
crops both during the spring and fall. For
IJE
sensitive to the weather, accurate weather

example, peach tree in full bloom can have
forecasting is essential considering the fact
their potential peach crop decimated by a
that a greater number of plane crashes
spring freeze. Orange groves can suffer
recorded the world over have weather
significant damage during frosts and
related causes. Just as turbulence and icing freezes, regardless of their timing.
are significant in flight hazards,
thunderstorms are a major problem for all 5.5 Utility companies
aircrafts because of severe turbulence due
to their updrafts and outflow boundaries, Electricity and gas companies rely on
icing due to the heavy precipitation, as weather forecasts to anticipate demand,
well as large hail, strong winds, and which can be strongly affected by the
lightening, all of which can cause severe weather. They use the quantity termed the
damage to an aircrafts in-flight. Volcanic degree-day to determine how strong of a
ash is also a significant problem for use there will be for heating (heating
aviation, as aircrafts can lose engine power degree day) or cooling (cooling degree
with ash clouds. On a day-to-day basis, day). These quantities are based on a daily
airliners are routed to take advantage of average temperature of 650F (180C).
the jet stream tailwind to improve fuel Cooler temperatures force heating degree-
efficiency. Aircrews are briefed prior to days (one per degree Fahrenheit), while
takeoff on the conditions to expect enroute warmer temperatures force cooling degree-
and at their destination. Additionally, days. In winter, severe cold weather can
airports often change which runway is cause a surge in demand as people turn up
being used to take advantage of a their heating. Similarly, in summer or dry
headwind. This reduces the distance season a surge in demand can be linked
with the increased use of air conditioning

ISSN: 2278-0181
systems in hot weather. By anticipating a So far, the accuracy of long range

surge in demand, utility companies can forecasting has been minimal, but the short
produce additional supplies of power or range forecasting has been of immense
natural gas before the price increases, or in help and advantage to the world at large
some circumstances, supplies are restricted today.
through the use of brown outs and
blackouts. REFERENCES
5.6 Private Sector Ackerman, S. A. and Knox, J. A. (2003).

Meteorology: Understanding the
Increasingly, private companies pay for
Atmosphere. Brooks/Cole USA, pp
weather forecasts tailored to their needs so
that they can increase their profits or avoid 362 – 379
large losses. For example, supermarket Aguado, E and Burt, J. E. (2001)
chains may change the stocks on their Understanding Weather and
shelves in anticipation of different, Climate, Second Edition. Prentice
consumer spending habits in different Hall, New Jersey, pp. 348 – 355.
weather conditions. Weather forecasts can
be used to invest in the commodity market, Barry, R. G. and Chorley, R. J. (1992).
such as futures in oranges, corn, soybeans Atmosphere, Weather and Climate,
and oil. Also, members of the public use Sixth edition, Routledge, New
knowledge of future weather conditions to York, pp. 175 – 184
determine what to put on, on a daily basis.
Houghton, J. T. (1986). The Physics Of
5.7 Military applications Atmosphere, Second Edition
RT
Similarly to the private sector, military Published by the Press Syndicate of
weather forecasters present weather the University of Cambridge, New
conditions to the war fighters, community. York, pp. 165 – 188.
IJE
Military weather forecasters provide pre-

flight weather briefs to pilots and provide Linacre, E and Geerts, B. (1997). Climates
real time resource protection services for and Weather Explained Routledge
military installations. London, pp. 321 – 345
6.0 CONCLUSION Lutgens, F. K. and TarBuck, E. J. (1989).

The Atmosphere: An Introduction
Weather forecasting is a complex and to Meteorology, Fourth edition.
challenging science that depends on the Prentice Hall, New Jersey, pp. 299
efficient interplay of weather observation, – 331.
data analysis by meteorologists and
computers, and rapid communication Miller, A. and Thompson, J. C. (1975).
systems. Meteorologists have achieved a Elements of Meteorology, Second
very respectable level of skill for short- Edition. A Bell and Howell,
range weather forecasting. Further Columbus, Ohio pp. 205 – 233.
improvement is expected with denser
Mohan, J. M. and Morgan, M. D. (1991).
surface and upper air observational
Meteorology: The Atmosphere and
networks, more precise numerical models
Science of Weather, Fourth edition.
of the atmosphere, larger and faster
Macmillan Ontario, pp. 356 – 381.
computers and more are to be realized.
However, continued international co- http://www.wikipedia.com/weatherforecast
operation is essential, for the atmosphere is ingmodels&applications
a continuous fluid that knows no political
boundaries.

Weather Forecasting Models Methods and Applications IJERTV2IS120198

Uploaded by

Weather Forecasting Models Methods and Applications IJERTV2IS120198

Uploaded by

International Journal of Engineering Research & Technology (IJERT)

Weather Forecasting Models, Methods and Applications

Iseh. A. J. * Woma. T. Y.1,2

ABSTRACT weather data, forecasting methods and

POexp–gz/RT. Computerized numerical growers depend heavily on accurate

IJERTV2IS120198 www.ijert.org 1945

be taken manually by a weather observer,

IJERTV2IS120198 www.ijert.org 1946

Greenwish Mean Time (0000GMT, of points where there are processed by a

pressure, and relative humidity up to an

IJERTV2IS120198 www.ijert.org 1947

atmosphere. Data assimilation proceeds by Referring to the work of Houghton (1986),

types of weather forecasting are based on

IJERTV2IS120198 www.ijert.org 1948

−g which are too slow to be important in short

At sea level, Z0 = 0. Thus, Meteorological bureaux now regularly

g is acceleration due to gravity. But of all these three types of numerical

IJERTV2IS120198 www.ijert.org 1949

7. ARPEGE – developed by the 6. High Resolution Limited Area

new understanding of the Physics of

IJERTV2IS120198 www.ijert.org 1950

3.0 FORECASTING PROBLEMS i. Imperfect data: The data of

IJERTV2IS120198 www.ijert.org 1951

ahead is the average value of past

IJERTV2IS120198 www.ijert.org 1952

4.3 Looking at the sky necessitating site-specific forecasts.

drop is rapid, a low-pressure system is airports is one example of the importance

IJERTV2IS120198 www.ijert.org 1953

IJERTV2IS120198 www.ijert.org 1954

sensitive to the weather, accurate weather

IJERTV2IS120198 www.ijert.org 1955

systems in hot weather. By anticipating a So far, the accuracy of long range

5.6 Private Sector Ackerman, S. A. and Knox, J. A. (2003).

Military weather forecasters provide pre-

6.0 CONCLUSION Lutgens, F. K. and TarBuck, E. J. (1989).

IJERTV2IS120198 www.ijert.org 1956

You might also like