Free Paper Samples

We have a large team of skilled writers to choose from to complete your order.


Attainment of precision in weather prediction is difficult. This could be attributed to deadly appearing storms turning out to be of less harm while the weaker ones devastate, precipitation falling abruptly and temperatures rising above expected levels. With the passage of time, technological and mathematical advancements have been invented and inculcated in weather prediction. This has had an improvement in the weather forecast. However, algorithms applied in weather forecasting lead to the production of static results despite the changing nature of weather. Methods such as observation of the sky were used in the past and are still in use at present in prediction of weather.















                        Air movement in the atmosphere is one of the causes leading to changes in weather. Models that utilize the grid system have been developed to study changes caused by this movement. Computers are used in working out mathematical calculation from the grid system using mathematical models. This leads to the production of equations that are used in predicting future weather patterns for a specific location. This paper is therefore aimed at identifying types of models and parameters applied in weather forecasting, advancements made in the forecast field, the difference between the present and past forecast methods and try to explain how calculation errors accumulate from the use of time periods in the prediction of future weather trends. Moreover, the paper will utilize the use of examples and illustrations in further projection the effects of accumulated errors.

Forecasting models

Two forecast models exist, the global model which is used to produce weather patterns for all parts of the earth and the regional model that covers specific parts of the earth. According to Kukkonen et al. (2012), the models use the grid system and Model Output Statistics (MOS) in study areas. The use of a grid system and MOS in identifying the exact predictions for the surface being analyzed enables the production of finer details for a particular point. The use of more grid reference points enhance precision (Kukkonen et al., 2012). MOS utilize data from observations in predicting future patterns. Once the data is fed into the model, computers are used to generate future trends by use of time steps. Previous short future predictions are used in computing patterns for later dates. Kukkonen et al. (2012) went further and stated that MOS is divided into short range (SR) that produces analysis in an interval of six hours and the long range (LR) that produces data in a twelve-hour interval.


Key Parameters in Weather Forecast

The MOS applies a number of weather elements in predicting future weather trends. According to Stensrud (2009), air temperature is one example and is measured in Fahrenheit at two meters above the ground. The speed and direction of the wind is another parameter. For wind speed, it is given in knots. Wind direction, on the other hand, is measured at heights of 10 meters above sea level and is given in tens of degrees. Precipitation is the other parameter. Under precipitation, rainfall, dew, snowfall, freezing rain and a mixture of drizzles and snow is studied. Cloud cover presents another parameter. Stensrud (2009) explained that the sky is studied to determine whether it is clear, has few clouds, has scattered clouds, broken clouds or overcast. Another parameter is haze and content of the air. Under this, atmospheric visibility and obstruction are studied.

Advances Made to Enhance Forecast Accuracy

A number of technological and algorithm inventions have been made to improving precision in the weather forecast. Improvements in radar technology and other technologies associated with weather sensing has enhanced accuracy and provision of early warnings in the face of an incoming weather change (Allen, Brown, Lewitus, & Sandifer, 2015). The invention of cheap sensors that provide computers with real-time information for numerous locations has enabled meteorologists to predict the weather for specific areas. Moreover, the new Doppler radar units are capable of peering through an incoming front and measure the speed of wind (Allen et al., 2015).

Improvements have also been experienced on computers used in weather forecasting. Computers that process data at a fast rate have been developed. With improvements in the number of sensors produced, these computers produce weather averages based on sensor information relayed from different locations. Allen et al. (2015) argued that availability and use of more sensors improve accuracy by reducing the focus area and feeding the computer with more weather elements.

The invention of the numerical weather prediction model presents another advancement. Day to day forecast has been boosted by the use of mathematical equations in the prediction of weather. Few changes in weather parameters occur between one to two days. Therefore, accuracy improvement has been witnessed particularly in the short-range weather forecast (Bauer, Thorpe, & Brunet, 2015).

Invention and improvements on the existing mathematical algorithm is another milestone in enhancing accuracy in weather forecasting. Algorithms such as the ID3 and the sliding window have gone through transformations. For instance, the Dynamic MRI is as a result of improvements on the sliding window algorithm. It providing real-time imaging (Sumbul, Santos, & Pauly, 2009). The images are less blurred and can be used to predict future weather patterns using immediate weather conditions.

Difference between past and present Forecasting Methods

There has been a shift in the approach to predicting the weather. In the past, around 650BC, skies were observed. In around 300BC, a calendar was developed, dividing the year into twenty-four festivals depending on the weather patterns associated with the periods (Allen et al., 2015). At present, the weather forecast is carried out using technology. Radar, satellite and sensor technology is used in observing the skies. Consequently, the calendar year is divided into four major seasons. Besides, unlike in the past, present weather prediction involves mathematical calculations using computers and pre-determined algorithms in predicting the weather. The change has enhanced accuracy in weather prediction.

The methods of relaying weather predictions have also evolved. Initially, information was relayed by texts or word of mouth. At present, digital methods are used to relay information on weather (Allen et al., 2015). For instance, the use of digital database forecasting offers automated services to users by sending email texts and alerts. Real-time systems detect echoes produced by radar and apply extrapolation thus giving hourly updates on weather changes. Use of Information technology in weather prediction is another difference that exists between past and present predictions. The internet has been used to relay weather forecasts and warnings to customers, and the general public in graphics and general formats. Geographical Information system and the Global positioning system have been recently used in capturing, storing and relaying weather predictions. According to Allen et al. (2015), this has been possible because of the integration of GIS and GPS systems with mobile communication devices such as cell phones. This presents a great shift from the past.

Accumulation of Input Variation and Processing Errors

Prediction on weather is done using time steps. Variables are fed into a computer and prediction made on future weather conditions. From the results obtained, Subsequent calculations are done to predict the weather for dates beyond the first time step. Presence of an error in the first time step will multiply and accumulate in subsequent calculations. For example, if it rained only for ten minutes in one hour and rain duration is expected to be constant for three hours, then the error will be calculated as

For the first hour, it rains for ten minutes, we will have,

1/6*100=16.7% of an hour of rain

If for the next three hours it rains for 20 minutes, we will have 

2/6*100=33.3% of an hour of rain

However, if we used the time step, we would have

3/6*100=50% of an hour of rain

Therefore the error is 50%-33.3%= 16.7%

              Other than the accumulation of errors as portrayed above, weather variables keep on changing thereby making it difficult to use a rigid mathematical formula in the prediction of future weather. This has made prediction difficult and inaccurate.


Weather forecasting has come a long way from sky observation by the Babylonians and Greeks to the invention of the calendar by the Chinese, to the present age of technology in observation, prediction, and relay of weather patterns. Mathematical models and algorithms have been applied in the prediction of future weather. Nonetheless, precision has not been achieved even with the use of supercomputers and Doppler radar technology. Instead, the accumulation of