If we can't predict the weather accurately for next week, you might ask yourself, why do we think we can make predictions for next season? Well, we can't predict the weather for next season, but under some conditions we can to say something useful about the climate for next season.
Weather versus climate prediction It is well known that weather forecasts are usually fairly accurate in terms of predicting the significant weather features for the coming 1 to 2 days. It is also known that the accuracy of weather forecasts decreases as the lead time increases to 3, 4, or 5 or more days. A forecast for 4 days into the future, for example, often needs to be revised as that day approaches, and in some cases the revision may be large. At lead times of 5 to 7 days there is only a small amount of accuracy in weather forecasts, and over 7 days there is nearly none. Why, then, is it possible to make useful forecasts for some regions for the coming three months, and sometimes even for the 3-month period following this? At seasonal lead times, there is no usable skill in forecasting on which day a place will have precipitation, storms, temperature extremes etc. This is consistent with the rapid drop-off of skill after several days, discussed above. However, there is nonetheless some skill in predicting anomalies in the seasonal average of the weather—i.e., anomalies of the climate. This skill is present regardless of the daily timing of the major weather events within the period. The total precipitation, for example, may be predicted to be higher than the climatological average due to a greater-than-normal expected frequency of a specific atmospheric circulation pattern that is conducive to rainfall at the location in question. Again, the timing of the rainfall events remains unknown. Forecasts of the likelihood of enhanced or suppressed rainfall or lower or higher temperatures than the average, over the course of a season have a level of accuracy that is far from perfect but noticeably above the level of random chance. This level of skill for seasonal averages or totals may be useful for sectors impacted by climate variability, such as energy production, agriculture, health, and others.
What’s the basis of forecasting?
Much of the skill in predicting departures from normal seasonal totals or averages, often associated with atmospheric circulation patterns, has its origin in the slowly changing conditions at the earth’s surface that can influence the climate. The most important surface condition affecting climate is the sea surface temperature (SST), and particularly the SST in the tropical zones. Other, usually less influential, surface conditions are soil wetness and snow cover. The feature of the surface conditions that gives them the ability to influence the average of the weather conditions over an extended future period is the slowness with which they can change, and therefore the extended period over which they can exert their consistent influence. When the SST is higher than normal, it usually remains that way for several months, and sometimes for as long as a year or more, such as during the El Nino or La Nina (i.e., the warm and cold phases of the ENSO – the El Nino/Southern Oscillation) episodes of the tropical Pacific SST. Similarly, when there is high soil wetness, or snow cover, it often takes at least several weeks for this situation to return to normal, because on each day the sun can only evaporate or melt a limited portion of the excess. When the soil is very dry, it may take 4 to 8 significant rainfall events to bring the soil wetness back to its normal, since the water from one heavy rainfall often runs off and may not replenish the soil wetness more than superficially. Anomalies of the SST are particularly slow to change because of the high heat capacity of water relative to the atmosphere, both because of its higher density and because its anomalies may extend to many tens of meters beneath the ocean surface. Slowness in the variations of SST implies that significant departures from normal (i.e., anomalies) of the currently observed SST are likely to persist into the coming several months. It also means that if SST anomalies can be predicted with some reliability (which is the case in certain important instances, to be explained below), then the climate that is dynamically associated with the SST anomalies can also be predicted somewhat reliably.
What is the difference between Accuracy and Skill?
Accuracy is the ability of a single forecast to predict the conditions correctly. Skill is the ability of forecasts to continually predict the conditions with some accuracy. In other words skill describes the ability to regularly produce accurate forecasts.
How skilful are the CSAG seasonal forecasts?
The CSAG forecast has only been operational for about 5 years. During this time its forecasts have been verified against observed rainfall for specific 3-month periods, for lead times of 1 month and 3 months.(that is forecasts produced 1 or 3 months before the forecast period) The skill has been calculated to produce a Forecast Index which indicates the skill on a scale of -10 (max negative skill) through 0 (no skill) to +10 (maximum skill). Negative skill means that the forecast was predicting almost the exact opposite of what was observed. Such forecasts may be useful after all! To answer the question: the skill of the forecasts range from -7.2 to + 7.6. They are colour coded in the table below to be able to differentiate the skill levels. However, we encourage the use of these forecasts only if the skill is higher than +2, and then to realise that a skill of 2-4 is only moderate. It will be seen that the forecasts are generally MORE skilful during the rainy seasons for the two areas tested. The skill for the two lead times varies considerably.
O N D
Should I use these forecasts in decision-making? One might wonder whether, and how, forecasts that have only modest levels of skill can be used beneficially in decision making. The answer to these very valid questions, firstly, is that they can indeed be used beneficially, when used very carefully. If used inappropriately, they may not be beneficial and in fact could be detrimental, at least in the short term. The economic benefits from using the forecasts properly should be expected to grow over an extended time frame, and to grow somewhat irregularly, rather than to appear immediately and to occur uniformly from one year to the next. This expectation follows from the fact that climate forecasts are expected to be incorrect some of the time, but to be correct more frequently than incorrect. (By "correct", we mean that the forecast verified at least in terms of the direction of the deviation from observed means) Because of the forecasts’ uncertainty, decisions should be made cautiously, but using criteria that are consistent over time, in accordance with the probability anomalies that are forecast. The cost of taking precautions must be weighed against the savings that the precautions would bring if the unwanted climate event occurred. These judgements can be made using the probabilities supplied in the forecasts. Because some businesses cannot necessarily count on existing for the long time frame needed to be sure to benefit from using climate forecasts, a major consideration is the short-term consequences of a loss during the coming individual season that would be caused by a forecast that does not verify.
If such a loss would result in an unacceptable probability of bankruptcy, then the long-term use of the climate forecasts would need to be modified accordingly. One strategy for doing this would be to act on the direction of the forecast but replace the given probabilities with more conservative probabilities—i.e. ones that more weakly differ from averages. A detailed cost-benefit analysis, which would differ greatly from one business to another, would yield decision recommendations, depending on all relevant potential rewards and penalties. Such an analysis is strongly encouraged, as it is a vital part of the task of climate forecast applications.