Weather Forecasting Models, Methods And Applications

2y ago
28 Views
2 Downloads
552.36 KB
12 Pages
Last View : 20d ago
Last Download : 2m ago
Upload by : Olive Grimm
Transcription

International Journal of Engineering Research & Technology (IJERT)ISSN: 2278-0181Vol. 2 Issue 12, December - 2013Weather Forecasting Models, Methods and ApplicationsAA1Iseh. A. J. * Woma. T. Y.1,21. 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.weather data, forecasting methods andapplications.1.0INTRODUCTIONModern society’s ever-increasing demandfor more accurate weather forecasts isevident to most people. The spectrum ofneeds for weather predictions ranges fromthe general public’s desire to know if forinstance, the weekend will permit anouting at the beach, or an organization’srally, or an outdoor wedding reception.Such diverse industries as airlines and fruitgrowers depend heavily on accurateweather forecasts to have an idea of whattheir next schedule of flight would appearto be or if the weather will be suitable forharvesting. In addition, in developedcountries, the designs of buildings, andmany industrial facilities rely heavily on asound knowledge of the atmosphere.IJERTABSTRACTWeather forecasting modelling is acomputerprogramthatprovidesmeteorological information for futuretimes at given locations. In modernforecasting models, Numerical WeatherPrediction is mostly applied and thismeans, “a set of simplified equations usedto calculate changes in atmosphericconditions”. The act of writing theseequations,imposingtheboundaryconditions and solving them using supercomputers, is known as numericalmodelling. An example of such equationsis the Hypsometric equation given as PI POexp–gz/RT. Computerized numericalmodels are designed for different intervalswhich are known as global models underwhich we have long range forecast andmedium range forecast, and regionalmodels under which we have the shortrange forecast. The methods includepersistence, climatologic, looking at thesky, use of barometer, nowcasting, use offorecastingmodels,analogueandensemble forecasting. Forecasting could beapplied in air traffic, severe weather alerts,marine, agriculture, utility companies,private sector and military application.Weather forecasting is a complex andchallenging science that depends on theefficient interplay of weather observation,data analysis by meteorologist andcomputers, and rapid communicationsystem.Key words:Weather, weatherprediction, forecast, forecasting models,IJERTV2IS120198Weather forecasting can be defined as theact of predicting future weather conditionsor an attempt to indicate the weatherconditions which are likely to occur.Weather forecasting is the application ofScience and Technology to predict thestate of the atmosphere for a future timeand a given location. Human beings haveattempted to predict the weatherinformally for millennia, and formallysince at least the nineteenth century.Weather forecasts are made by collectingqualitative data about the current state ofthe atmosphere and using scientificunderstanding of atmospheric processes towww.ijert.org1945

International Journal of Engineering Research & Technology (IJERT)ISSN: 2278-0181Vol. 2 Issue 12, December - 2013project how the atmosphere will evolvewithin the next few hours.HOWMODELSCREATEFORECASTS2.1Data collectionSince invention of the first weatherinstruments in the seventeenth centuryweather observation has undergoneconsiderablerefinement.Densermonitoring networks, more sophisticatedinstruments and communication systems,and better-trained weather observers, haveproduced an increasingly detailed, reliableand representative record of weather andclimate. In weather forecasting, datacollection has been divided into twocategories namely:i.ii.Surface weather observationsUpper-air weather observations.Surface Weather ObservationsSurface weather observations are thefundamental data used for safety as well asclimatological reasons to forecast weatherand issue warnings worldwide. They canbe taken manually by a weather observer,by computer through observers to augmentthe otherwise automated weather station.IJERTOnce, an all-human endeavour basedmainly upon changes in barometricpressure, current weather conditions andsky conditions, forecast models are nowused to determine future conditions. Amodel, in this context, is a computerprogram that produces meteorologicalinformation for future times at givenpositions and altitudes. The horizontaldomain of a model is either global,covering the entire earth, or regional,covering only part of the earth. Regionalmodels also are known as limited areamodels. Human input is still required topick the best possible forecast model tobase the forecast upon, which involvespattern recognition skills, knowledge ofmodel performance and knowledge ofmodel biases. The chaotic nature of theatmosphere, error involved in measuringthe initial conditions, an incompleteunderstanding of atmospheric processesmean that forecast become less accurate asthe difference in current time and the timefor which the forecast is being madeincreases.2.0There are a variety of end users to weatherforecasts. Weather warnings are importantforecasts because they are used to protectlife and property. Forecasts based ontemperatureandprecipitationareimportant to agriculture, and therefore tocommodity traders within stock markets.Temperature forecasts are also used byutility companies to estimate demand overcoming days. On an everyday basis peopleuse weather forecasts to determine what towear on a given day. Since in recent timein Uyo – Nigeria for example, outdooractivities are severely curtailed by heavyrains, forecasts can be used to planactivities around these events, and to planahead and survive them.IJERTV2IS120198Referring to Lutgens and TarBuck (1989),a vast network of weather stations requiredto produce a weather chart will encompassenough to be useful for short-rangeforecasts. On a global scale, the Worldmeteorologicalorganization,whichconsists of over 130 nations, is responsiblefor gathering the needed data andproducing some general prognostic charts.Surfaceweatherobservationsofatmospheric pressure, temperature, windspeedanddirection,humidity,precipitations are made near the earth’ssurface by trained observers or automaticweatherstations.TheWorldmeteorological Organization acts tostandardize the instrumentation, observingpractices and timing of those observationsworldwide. By international agreement,the regular synoptic observations are madeevery six hours beginning at midnightwww.ijert.org1946

International Journal of Engineering Research & Technology (IJERT)ISSN: 2278-0181Vol. 2 Issue 12, December - 2013Greenwish Mean Time (0000GMT,0600GMT, 1200GMT and 1800GMT)each day. In addition, observations foraviation purposes are made at manyairports every hour, or more often if theweather is changing rapidly. Specializedobservations may also be made undercertainconditionsforagricultural,industrial, research, or other purposes(Miller and Thompson, 1975).2.1.1DataassimilationandanalysisIn order to do their work, most numericalmodels look at the atmosphere as a seriesof boxes. In the middle of each box is apoint for which the model actuallycalculates weather variables and makesforecasts. The result of this threedimensional boxing up of the atmosphereis known as the grid; the point in themiddle is the grid point, and the distancebetween one point and another is called thegrid spacing (Ackerman and Knox, 2003).IJERTUpper-air weather observationsMeasurements of temperature, humidityand wind above the surface are found bylaunchingradiosondesonweatherballoons. Radiosonde is a unit for use inweather balloons that measures variousatmospheric parameters and transits themto a fixed receiver. Radiosondes mayoperate at a radio frequency of 403MHz or1680MHz and both types may be adjustedslightly higher or lower as required.Mohan and Morgan (1991) states that, theinstrument transmits to the ground stationvertical profiles of air temperature,pressure, and relative humidity up to analtitude of about 30km. In addition, windsat various levels are computed by trackingthe balloons with a radio direction findingantenna.of points where there are processed by aCentral weather analysis organization.Three locations have been designated bythe World Meteorological Organization asWorld Meteorological Centers, these arelocated in Melbourne, Australia; Moscow,Russia; and Washington, D.C. U.S.A. Inaddition, most countries maintain nationalcenters, where the basic weather needs ofthe domestic economy are met. In Nigeria,the National Meteorological Center islocated in Abuja.Upper-air weather data are also obtainedby aircraft, dropwind sondes, radar, andsatellites. Increasingly, data from weathersatellites are being used because of theiralmost global coverage. Although theirvisible light images are very useful forforecasters to see development of clouds,little of this information can be used bynumerical weather prediction models. Theinfrared data however, can be used as itgives information on the temperature atthe surface and cloud tops. Individualclouds can also be tracked from one timeto the next to provide information on winddirection and strength at the cloudssteering level.Miller and Thompson (1975) agrees that,the observations are collected at a numberIJERTV2IS120198Grid point models of the atmosphere canget fussy when the data in the initialconditions is not obtained at exactly thelocation of the grid point. Also, the processof creating an evenly spaced data set fromirregularly spaced observations is calledinterpolation.Ackerman and Knox (2003) then say that,the multiple jobs of interpolating andsmoothing the data for use in numericalmodels are collectively called dataassimilation.During the data assimilation process,information gained from the observationsis used in conjunction with a numericalmodel’s most recent forecast for the timethat observations were made, since thiscontains information from previousobservations. This is used to produce athree-dimensional representation of thetemperature, moisture and wind called ameteorological analysis. This is the modelsestimate of the current state of thewww.ijert.org1947

International Journal of Engineering Research & Technology (IJERT)ISSN: 2278-0181Vol. 2 Issue 12, December - 2013atmosphere. Data assimilation proceeds byanalysis cycles. In each analysis cycle,observations of the current (and possibly,past) state of a system are combined withthe result from and mathematical model(the forecast) to produce an analysis,which is considered as “the best” estimateof the current state of the system. This iscalled the analysis step. Essentially, theanalysis step tries to balance theuncertainty in the data and in the forecast.The model is then advanced in time and itsresult becomes the forecast in the nextanalysis cycle.2.1.2Numerical weather predictionThere are numerous equations employed inthis work of forecasting models one ofwhich is the hypsometric equation whichcan be derived from the hydrostaticequation written asdP According to Lutgens and TarBuck (1989),the word “numerical” is misleading, for alltypes of weather forecasting are based onsome quantitative data and therefore couldfit under this heading. Numerical weatherprediction is based on the fact that thegases of the atmosphere obey a number ofknown physical principles. Ideally, thesephysical laws can be used to predict thefuture state of the atmosphere, given thecurrent conditions. This situation isanalogues to predicting future position ofthe moon based on physical laws and theknowledge of its current position. Still, thelarge number of variables that must beincluded when considering the dynamicatmosphere makes this task extremelydifficult.Manipulating the huge data sets andperforming the complex calculationsnecessary to do this (weather prediction)on resolution fine enough to make theresult useful requires the use of some ofthe most powerful supercomputers.IJERTV2IS120198-PgRTdZ-(1)We can simplify equation (1) by dividingthrough with P to haveIJERTLinacre and Geerts (1997) defineNumerical Weather prediction (NWP) as asimplified set of equations called theprimitive equation used to calculatechanges of conditions. Modern weatherforecasting relies heavily on numericalweather prediction.Referring to the work of Houghton (1986),the task of writing the equations and theboundary conditions in a suitable form andthen of solving them with high speeddigital computers is known as numericalmodeling. By comparing the behaviour ofthe model with that of the real atmosphere,the validity of the procedures employed bythe model is tested. The most importantapplication of numerical modeling is thedevelopment of methods sufficientlyreliable and sufficiently fast to be used inroutine weather forecasting.dPP g RTdZ-(2)Integrating the right hand side of equation(2) from P0 to P1 and the left hand sidefrom Z0 to Z1, we have:P 1 dPP0 PZ1ZRT0 g dZ CLet C 0InPP1P0 g ZRT InP1 – InP0 g RT g In (P1/P0) Z1Z0RT(Z1 – Z0)(Z1 – Z0) -(3)Taking the exponent of both sides ofequation (3), we have:𝑒 In (P 1P1 P0www.ijert.orgP0) g𝑒[ g𝑒[RT (Z 1 – Z 0 )]RT (Z 1 – Z 0 )]1948

International Journal of Engineering Research & Technology (IJERT)ISSN: 2278-0181Vol. 2 Issue 12, December - 2013 P1 P0𝑒 [ gwhich are too slow to be important in shortrange weather forecasting.RT (Z 1 – Z 0 )]At sea level, Z0 0. Thus,P1 –g Z 1P0𝑒 [RT ]-(4)Equation (4) is the Hypsometric equation,which gives the pressure PI of theatmosphere at a particular height Z1,Where P0 is the pressure at the sea level,g is acceleration due to gravity.R is the molar gas constant andT is the temperature at the surfaceBut of all these three types of numericalweather prediction, Mohan and Morgan(1991) say that actually, when viewed withthe objectivity of statistical analysis, shortrange weather forecasting is surprisinglymore accurate probably because of theshort range of time within which thisforecast is made for.The most modern models of weather dataprocessing systems for the two categoriesof numerical models are as follows:IJERTMohan and Morgan (1991) agree withBarry and Chorley (1992) that, somecomputerized numerical models of theatmosphere are designed to operate overdifferent spatial scales depending on theforecast range. For medium range forecasts(up to 10 days), observational data are fedinto the computer from all over the globe,since within that forecast range a weathersystem may travel long distances. On theother hand, for short-range forecasts (up to3 days), the model utilizes data drawnfrom a more restricted region of the globe.Compared to a global model, a regionalmodel offers the advantage of greaterresolution of data over a smaller area ofinterest.Meteorological bureaux now regularlyprovide seasonal outlook, with an accuracynotablyenhancedbyincreasedunderstanding of the relevance of thesouthern oscillation, indicated by the seasurface temperatures, the strength of thetrade winds, the location of areas of dayconvention across the tropical PacificOcean, and the depth of themocline.As it is, there are basically three types ofnumerical weather prediction modelsnamely: short-range forecasts, mediumrange forecast (as briefly discussed byMohan and Morgan, 1991; and Barry andChorley, 1992 above), and long rangeforecast.Thus, Linacre and Geerts(1997) describeslong range forecasts as descriptiveforecasts which are made for times of tento thirty days or one to four months (aseasonal outlook). Also, use is made of adynamical numerical weather predictionmodel which allows for oceanic processes,IJERTV2IS120198Global models: Some of the better-knownglobal numerical models are:1.Global Forecast System (GFS) –DevelopedbytheNationalOrganization for the Atmosphere inAmerica. Output is freely available.2.NOGAPS – Developed by the USNavy to compare with the GFS3.Global Environmental Multi-scaleModel (GEM) – Developed by themeteorological service of Canada4.European Centre for MediumRangeWeatherForecasts(ECMWF) – a model run by theEuropeans with limited availability5.UKMO – Developed by the UnitedKingdom Meteorological Office.Limited availability, but is handcorrectedbyprofessionalforecasters.6.GME – developed by the GermanWeather Servicewww.ijert.org1949

International Journal of Engineering Research & Technology (IJERT)ISSN: 2278-0181Vol. 2 Issue 12, December - 20137.ARPEGE – developed by theFrench Weather Service, MeteoFrance.8.Intermediate General Circulationmodel (IGCM) – developed bymembers of the Department ofMeteorology at the University ofReading.Regional models: Some of the betterknown regional numerical models are:TheWeatherResearchandForecasting (WRF) Model wasdeveloped co-operatively by NCEPand the meteorological researchcommunity. WRF has severalconfigurations including:a.b.2.3.4.5.IJERTV2IS120198WRF – NMM: The ERFNon-hydrostatic MesoscaleModel is the primary shorttermweatherforecastmodel for the United State,High Resolution Limited AreaModel (HIRLAM)7.GEM–LAM–GlobalEnvironmental Multi-scale LimitedArea Model8.Aladin: The high resolution pedoperated by several European andNorth African countries under theleadership of Meteo-France.9.COSMO: The COSMO Model,formerly known as LM, aLMD orLAMI, is a limited area nonhydrostaticmodeldevelopedwithin the framework of theconsortiumforsmallscalemodelling (Germany, om.According to Linacre and Geerts (1997),the advantage of numerical weatherprediction is that it avoids errors of humanjudgment in deriving the prognosis, andcan be steadily improved by enlarging theamount and reliability of input data, bynew understanding of the Physics ofatmospheric change, and by faster, largercomputers.IJERT1.6.AR-WRF:AdvancedResearch WRF developedprimarily at the United StateNationalCenterforAtmosphericResearch(NCAR)The North American MesocaleModel (NAM)Colorado State University fornumericalsimulationsofatmospheric meteorology and otherenvironmentalphenomenaonscales from metres to hundreds ofkilometresMMS – The fifth Generationmesoscale modelThe Advanced Region PredictionSystem (ARPS) – developed at theUniversity of Oklahoma. It is acomprehensive multi-scale nonhydrostaticsimulationandprediction system that can be usedforregionalscaleweatherprediction up to the tornadscalesimulation and prediction.2.1.3Model Output Post ProcessingThe raw output is often modified beforebeing presented as the forecast. This canbe in the form of statistical techniques toremove known biases (a term used todescribe a tendency or preference towardsa particular perspective, ideology or result)in the model, or its adjustment to take intoaccount consensus among other numericalweather forecast. MOS or Model OutputStatistics is a technique used to interpretnumerical model output and produce sitespecific guidance. This guidance ispresented in coded numerical form, andcan be obtained for nearly all NationalWeather Service reporting stations.www.ijert.org1950

International Journal of Engineering Research & Technology (IJERT)ISSN: 2278-0181Vol. 2 Issue 12, December - 20133.0i.FORECASTING PROBLEMSAn important goal of all scientificendeavour is to make accurate predictions.The physicist or chemist who conducts anexperiment in the laboratory does so in thehope of discovering certain fundamentalprinciples that can be used to predict theoutcome of other experiments based onthose principles. In fact, most of the lawsof science are merely very accuratepredictions concerning the outcome ofcertain kinds of experiments. But fewphysical scientists are faced with morecomplexorchallengingpredictionproblems than the meteorologist.ii.IJERTIn the first place, the meteorologicallaboratory covers the entire globe, so thateven the problem of measuring the presentstate of the atmosphere is tremendous.Furthermore, the surface of the earth is anirregular combination of land and water,each responding in a different way to theenergy source – the sun. Then, too, theatmosphere itself is a mixture of gaseous,liquid, and solid constituents, many ofwhich affect the energy balance of theearth, one of them, water, is continuallychanging its state. Also, the circulations ofthe atmosphere range in size fromextremely large ones, which may persistfor weeks or months, to minute whirls,with life spans of only a few seconds.According to Miller and Thompson (1975)and Ayado and Burt (2001) the problem offorecasting then, involves an attempt toobserve, analyze and predict the manyinterrelationships between the solar energysource, the physical feature of the earth,and the properties and motions of theatmosphere. This is the basis on whichweather forecasts still go wrong today.Ackerman and Knox (2003) points outreasons why forecasts still go wrong todayby stating that the limitations whichdirectly relates to today’s numericalforecast models are as follows:IJERTV2IS120198www.ijert.orgImperfect data: The data oftoday’s numerical models stillincludes a large helping ofradiosonde observations. However,the number of radiosonde sites inthe World over has actuallydeclined over the past few decades.Developed countries in the worldtoday, spend more money inlaunching weather satellites thanfor boring weather balloons.Satellite data are global in average,but researchers in data assimilationare still trying to figure out howthis data can be “digested” properlyby the models. In addition,important meteorological featuresstill evade detection, especiallyover the oceans. The model resultsare only as good as the data in itsinitial conditions.Faulty “vision” and “fudges”:Today’s forecasts also involve aninevitabletrade-offbetweenhorizontal resolution and the lengthof the forecast. This is because fineresolution means lots of point atwhich to make calculations. Thisrequires a lot of computer time. Aforecast well into the future alsorequires millions or billions morecalculations. If fine resolution iscombined with a long rangeforecast, the task would choke thefastest supercomputers today. Onewould not get forecasts for weeks.Future improvement in computingwill help speed things up.In the meantime, however, somemodels are still not able to pick or“see” small-scale phenomena suchaclouds,raindrops,andsnowflakes. To compensate for thisfuzzy “vision” of models, thecomputer code includes crudeapproximations of what is notbeing seen. These are calledparameterizations. Even thoughmuch science goes into them, theseapproximations are nowhere closeto capturing the complicated reality1951

International Journal of Engineering Research & Technology (IJERT)ISSN: 2278-0181Vol. 2 Issue 12, December - 2013of the phenomena. This is because;the smallest scale phenomena areoften the most daunting tounderstand. Therefore, it is not aninsult to meteorologists’ abilities tosay that parameterizations are“fudges” of the actual phenomena.iii.Chaos: It will be surprising to notethat, even if a supercomputer whichcoulddoquadrillionsofcalculations each second were to beinvented, no better forecastingresult would still be gotten. Bruteforcenumericalweatherforecasting with extremely fineresolution has its limits.Persistence forecasts are used by localforecasters in determining such events asthe time of the arrival of a thunderstormthat is moving toward their region.Persistence forecasts do not account forchanges that might occur in the intensity orin the path of a weather system, and theydo not predict the formation. Because ofthese limitations and the rapidity withwhich weather system change in mostgeographical regions, persistence forecastsbreak down after twelve hours, or a day atmost.4.2Climatology forecastingWhereas persistence forecasting is mostaccurate over short periods (before factorsfor change have had time to operate), thebest estimate of the weather a long timeahead is the average value of pastmeasurements there at that time of day andyear.IJERTThe reason for these limits is acurious property of complex,evolvingsystemsliketheatmosphere. It is called “Sensitivedependence on initial conditions”,and is a hallmark of what ispopularly known as chaos theory.Chaos in the atmosphere does notmean that everything is a mess;instead, it means that theatmosphere both in real life and ina computer model may read verydifferently to initial conditions thatare only slightly different.when it is steady state, such as during thesummer season in the tropics. This methodof forecasting strongly depends upon thepresence of a stagnant weather pattern. Itcan be useful in both short-range forecastsand long-range forecasts.Because we do not know theatmospheric conditions perfectly atany time, chaos means that theresemblance between a model’sforecast and reality will be less andless with each passing day.4.0METHODS OF WEATHERFORECASTINGOf course for weather forecast to existthere must be methods on which it is done.These methods are as follows:4.1Persistence forecastingClimatology forecast relies on theobservation that weather for a particularday at a location does not change muchfrom one year to the next. As a result, along term average of weather on a certainday or month should be a good guess asthe weather for that day or month. Themost obvious climatology forecast in thispart of the world (Nigeria) is, “Cold inDecember, warm in July (the popular Julybreak)”. One does not need to be ameteorologist to make that forecast.Today’s numerical forecast methods stilluse climatological statistics as a “realitycheck”. There make sure that the computermodels are not going off the deep end,climatologically speaking.Persistence forecasting is the easiestmethod of forecasting which assumes acontinuation of the present. It relies upontoday’s conditions to forecast the weatherIJERTV2IS120198www.ijert.org1952

International Journal of Engineering Research & Technology (IJERT)ISSN: 2278-0181Vol. 2 Issue 12, December - 20134.3Looking at the skyAlong with pressure tendency, use of thesky condition is one of more importantweather parameters that can be used toforecast weather in mountainous areas.Thickening of cloud cover or the invasionof a higher cloud deck is indicative of rainin the near future. Morning fog portendsfair conditions, as rainy conditions arepreceded by wind or clouds, which preventfog formation. The approach of a line ofthunderstorm could indicate the approachof a cold front. Cloud free skies areindicative of fair weather for the nearfuture. The use of sky cover in weatherprediction has led to various weather loreover the centuries.4.4Use of a barometerSeveral countries including Nigeria haverecently developed integrated satellite andradar systems to provide information onthe horizontal and vertical extent ofthunderstorms, for example. Such data are,supplemented by networks of automaticweather stations that measure wind,temperature and humidity.Nowcasting methods use highly automatedcomputers and image analysis systems tointegrate data from a variety of sourcesrapidly. Interpretation of the data displaysrequires skilled personnel and or extensivesoftwaretoprovideappropriateinformation. The prompt forecasting ofwind shear and downburst hazards atairports is one example of the importanceof nowcasting procedures.IJERTUsing barometric pressure and the pressuretendency (xthe change of pressure overtime) has been used in forecasting sincethe late 19th century. The larger the changein pressure, especially, if more than2.54mmHg, the larger the change inweather can be expected. If the pressuredrop is rapid, a low-pressure system isapproaching, and there is a greater chanceof rain. Rapid pressure rises are associatedwith improving weather conditions, suchas clearing inthiscategoryarethunderstorms, gust fronts, tornadoes, highwinds especially along coasts, over lakesand mountains, heavy snow and freezingprecipitation. The development of radarnetworks, new instruments and high speedcommunication links has provided a meansof issuing warnings of such phenomena.4.5NowcastingThe forecasting of the weather within thenext six hours is often referred to asnowcasting. In this time range, it ispossible to forecast smaller features suchas individual showers and thunderstormswith reasonable accuracy, as well as otherfeatures too small to be resolved by acomputer model. A human given the latestradar, satellite and observational data willbe able to make a better analysis of thesmall scale features present and so will beable to make a more accurate forecast forthe following few hours.Severe weather is typically short-lived(less than two hours) and, due to itsmesoscale character (less than one hundredkilometers), it affects local/regional areasIJERTV2IS1201984.6Use of Forecasting ModelsIn the past, the human forecasters wereresponsible for generating the entireweather forecast based upon availableobservation. Today, human input isgenerally confined to choosing a modelbased on various parameters, such asmodel biases and performance. Using aconsensus of forecast models, as well asensemble members of the various models,can help reduce forecast error. However,regardless how small the average errorbecomes with any individual system, largeerrors within any particular piece ofguidance are still possible on any givenmodel run. Humans can use knowledge oflocal effects, which may be too small insize to be resolved by the model to addinformation to the forecast.www.ijert.org1953

International Journal of Engineering Research & Technology (IJERT)ISSN: 2278-0181Vol. 2 Issue 12, December - 20134.7Analogue ForecastingThe analogue method is a complex way ofmaking a forecast, requiring the forecasterto remember a previous weather eventwhich is expected to be mimicked by anupcoming event. The analogue forecaster’stask is to locate the date in history whenthe weather is a perfect match, oranalogue, to today’s weather. Then theforecast for tomorrow is simple –

Weather forecasting is a complex and challenging science that depends on the efficient interplay of weather observation, data analysis by meteorologist and computers, and rapid communication system. Key words: Weather, weather prediction, forecast, forecasting models, weather data, forecasting methods and applications.Cited by: 4Page Count: 12File Size: 552KBAuthor: Iseh. A. J., Woma. T. Y.

Related Documents:

Weather instruments are used to measure and record the weather. Weather instruments can be found in weather stations on land. The Met Office has hundreds of weather stations all over the UK. Weather instruments are also found at sea. They are found on some ships, but mainly on weather buoys designed to monitor weather and sea conditions.

Unit 2 Weather 5 LESSON 1 Today’s Weather BIG IDEAS Weather affects the way we live, what we eat and wear and how we feel. We can describe weather conditions by using mathematics. LESSON 2 What Makes Weather? BIG IDEAS The sun heating the earth and its atmosphere causes the weather. We feel weather as wind, heat or cold, and humidity in the form of rain,

Winter Weather Safety Know Your Risk, Take Action, Be a Force of Nature weather.gov/safety. Building a Weather-Ready Nation // 2 NATIONAL WEATHER SERVICE Winter Weather Hazards . Keep an eye out for debris, downed power lines, and tree branches weather.gov/safety. Building a Weather-Rea

Weather instruments Weather data Working definition of weather . 4.1-3 PROCEDURES 1. Have the students work in small groups to make concept maps about weather. Have them Include different kinds of weather, tools used to measure weather, and any other ideas they have about weather.

Although forecasting is a key business function, many organizations do not have a dedicated forecasting staff, or they may only have a small team. Therefore, a large degree of automation may be required to complete the forecasting process in the time available during each forecasting and planning cycle.

ects in business forecasting. Now they have joined forces to write a new textbook: Principles of Business Forecasting (PoBF; Ord & Fildes, 2013), a 506-page tome full of forecasting wisdom. Coverage and Sequencing PoBF follows a commonsense order, starting out with chapters on the why, how, and basic tools of forecasting.

Undoubtedly, this research will enrich greatly the study on forecasting techniques for apparel sales and it is helpful to identify and select benchmark forecasting techniques for different data patterns. 2. Methodology for forecasting performance comparison This research will investigate the performances of different types of forecasting techniques

2016 h Autodesk d 3D designs onfidence us sign data. Sp maximize pr op, cloud, an Prev AutoCAD so with robust t ing TrustedD eed detailing oductivity. C d mobile solu iew ftware, one ools that can WG techno and docume onnect your w tions. AutoC Guide of the world’ produce alm logy, the ori ntation work workflow and AD delivers t s leading CAD ost any shap ginal and mo with tools b .