The Role of Weather in Risk Management For the Market Technician s Association October 15, 2013 RIA PERSAD President StatWeather rpersad@statweather.com
Predictive Analytics Consumer Behavior Astronomy Climate Stock Prices Techniques from one discipline can be applied to another in the quest to accurately predict future behaviors.
Topics How Weather Forecasts are Made Weather Models and Their Interpretation Long-Range Weather Forecasting and Probabilistic Forecasting for Risk Management The Accuracy of Weather Forecasts Climate Change and Extreme Weather
Conventional Method of Weather Prediction: Dynamical Models Numerical Weather Prediction : government models (GFS, European, etc) for short- and medium-range forecasts The calculation of tomorrow's weather will have some inaccuracy. To get the weather for the day after tomorrow, input tomorrow's weather (which has some error) into the model, and you will get the next day. Each day you iterate, the errors are compounded. This way of forecasting breaks down considerably after 8 days.
Long-Term Forecasting Conventional methods do not use historical data. Using historical climate data (for example, 120 years of climate history), patterns can be found and mined from the data which enables the calculation of certain patterns to occur again in the future. Powerful computer programs can analyze hundreds of thousands of patterns and give probabilities of future weather patterns occurring given the recent conditions. Analysis of the past teaches us about the future and what can happen even years into the future. Forecasts can be presented with risk probabilities.
Longer Lead Times How can weather be predicted months or years in advance? Climatology e.g., El Niño and long-term atmospheric indicators Pattern Recognition mining historical data Analog Methods Neural Networks Statistical Correlations
Risk in Forecasting Probability of event occurring Margin of error Forecaster accuracy
A Probabilistic Forecast 1-15 day ahead weather forecasts and 16-90 day ahead weather forecasts for timing and targeting extreme events (heat waves, cold spells, frost and freeze, dry spells, peaks for demand response) Probabilities, confidence, and margins of error for risk management
How Good Can a 16-90 Day Forecast Be? Every major heat wave in the U.S. during Summer 2013 was able to be targeted with at least 30 days notice. Forecasting by using an effective statistical algorithm works!
Risk Management Tools Meticulous daily tracking and reporting of the accuracy of weather forecasters by city, state, region, and country and comparing different forecasters to find consensus.
Subjectivity vs. Objectivity Expert computer programs can optimize and automate the weather forecasting process, continually updating and improving the process. An objective computer system is required for the calculation of risk. Meteorologist vs. Computer Program: https://www.youtube.com/watch?v=yilq9c_7y7i
Exactly How Accurate are Weather Forecasts? StatWeather scientists have been tracking the accuracy of weather forecasts in the United States for 13 years. In 1-5 day forecasts, the error among most forecasters is currently about +/-2.5 deg F, or an error range of 5 deg F. In 6-10 day forecasts, the error is about 4 deg F, or an error range of +/-8 deg F. These are averages. A particular day can be as much as 18 deg F off. There can be much forecaster volatility from day to day. Using Numerical Weather Prediction / government models, the accuracy beyond 10 days is comparable with using climate averages. (GFS: ~7 days, Euro: ~7.5 days) Using Statistical Pattern Recognition, accuracy can extend weeks and even months ahead. Example: Energy Risk reported StatWeather's 6-month ahead statistical forecasts to have an accuracy of 74% as compared with more traditional forecasting methods (NOAA) which have 20%-35% accuracy. For more information, read Ria Persad's articles on Energy Central: www.energyblogs.com/weather
Season-Ahead Accuracy The actual outcome is at left; StatWeather's prediction is center; the U.S. Government prediction is at right. Red color is warmer than 30-year average; blue color is colder than 30-year average. The U.S. Government uses Numerical Weather Prediction, which is not very accurate for long term patterns. Go to our website, www.statweather.com for more comparisons.
How the climate is changing Precipitation changes trending over the coming decades
Climate Change Update 2013-2014 Temperature changes Recent 10-Year Record DR. HANS VON STORCH, Lead Author on the IPCC: So far, no one has been able to provide a compelling answer to why climate change seems to be taking a break. We're facing a puzzle. Recent CO2 emissions have actually risen even more steeply than we feared. As a result, according to most climate models, we should have seen temperatures rise by around 0.25 degrees Celsius (0.45 degrees Fahrenheit) over the past 10 years. That hasn't happened. In fact, the increase over the last 15 years was very close to zero. This is a serious scientific problem that the Intergovernmental Panel on Climate Change (IPCC) will have to confront when it presents its next Assessment Report late next year..there are two conceivable explanations -- and neither is very pleasant for us. 1) because greenhouse gases, especially CO 2, have less of an effect than we have assumed, 2) in our simulations, we have underestimated how much the climate fluctuates owing to natural causes.
Five Differing Viewpoints on Climate Change
Extreme Weather and Climate Change In the United States, 2012 has been the hottest year out of the last 120 years of record-keeping Note: The 2012 drought is not as bad as the droughts in the 1930 s and 1950 s. It has been a year of extremes and even a decade of extremes ---some places have been the wettest they ve ever been in 120 years, other places the driest; some places the hottest, other places the coldest. We are seeing GREAT volatility the upcoming Fall and Winter. Some attribute this to natural cycles; others attribute this to man-made causes. Risk Management is more important than ever before.
Extreme Weather Impacts New pressures from extreme weather conditions, an evolving regulatory landscape and the increasing convergence of our electricity, water and telecommunications services will challenge a long-established industry to rapidly evolve across all aspects of it core operations generation, transmission and distribution. Utilities infrastructure across the country faces challenges associated with extreme weather, population shifts and aging infrastructure. Addressing these challenges requires investment In an era fraught with uncertainty and utilities will benefit from working with consultants. The brief period of extremely cold weather in January created a confluence of many gas buyers, few gas sellers and scarce pipeline capacity for the Northeast. As a result, the laws of supply and demand prevailed, and gas prices rose sharply, causing power prices to spike in tandem. This costly scenario emphasizes the need for thorough planning.
The Data Analytics Curve
Additional Information For more information, go to www.statweather.com or email service@statweather.com Thank you for joining us!