Forecast/forecasting long-term

Attempts to forecast long-term future technology in any area of science is an impossible task left to individuals with relatively high egos. However, it is possible to state some very general trends which the polyethylene industry may follow over the course of the next 10-20 years based on the present day status of ethylene polymerization catalyst technology. 

The structural time series analysis methods also referred to as state-space methods [HAR 86, COM 07] have been used more and more for modeling the aggregate number of fatalities at national level [DUP 07], The approach that the authors adopt is innovative, as usually such analyses are led on an annual basis - in order to explain and forecast long-term changes in the aggregate number of fatalities at national level [LAS 01]). On the contrary, short-term changes can only be modeled on an infra-annual basis similar but uncompleted approaches were taken on a quarterly basis - without the inclusion of exogenous variables [COM 2007] and on a monthly basis - without the inclusion of economic variables [BER 13]. 

Table 4. Long-Term Rubber Consumption Forecast by Area, 10 t...

A particular problem is the number of events that should be simulated before the results are stabilized about a mean value. This problem is comparable to the question of how many runs are required to simulate a Gaussian distribution within a certain precision. Experience shows that at least 1000 sample arrivals should be simulated to obtain reliable simulation results. The sample load (samples/day) therefore determines the time horizon of the simulation, which for low sample loads may be as long as several years. It means also that in practice many laboratories never reach a stationary state which makes forecasting difficult. However, one may assume that on the average the best long term decision will also be the best in the short run. One should be careful to tune a simulator based on results obtained before equilibrium is reached. 

The values of raw materials and products can be found in trade journals such as Chemical Marketing Reporter (published by Schnell Publishing Company), European Chemical News and Asian Chemical News (published by Reed Business Information). However, the values reported in such sources will be subject to short-term fluctuations, and long-term forecasts will be required for investment analysis. 

The specific consumptions shown are forecasts for the period 2010 to 2020 based on the prototypes available today. In the long term, an increasing share of hybrid-vehicle... 

Pharmaceutical companies have to be continuously vigilant and forecast the future directions of drug development and regulatory requirements. They have to use their core technical competencies to deliver a pipeline of products to remain competitive and profitable in the long term. 

Solomon, S and D. L. Albritton, Time-Dependent Ozone Depletion Potentials for Short- and Long-Term Forecasts, Nature, 357, 33-37 (1992). 

Since this estimated share pattern was derived mainly from projection of trends (particularly long term trends), it seems appropriate to focus on oil and speculate as to how possible future events might alter its forecast future role. Events related to pollution control tend to indicate increases in petroleum demand. The use of lead free gasoline, for instance, requires additional refinery processing, which in turn consumes more petroleum fuel. Increasingly tighter controls on sulfur dioxide emissions from thermal-electric plants will cause a shift from coal to low sulfur fuel oil if there is no economic flue-gas desulfurization to cope with coals sulfur content. 

---