Figure 4 Winds blow from the mountains down to the coast around Boston, driving temperatures up. Image of Boston Harbor taken Aug. 2, 2005 by IKONOS, courtesy of GeoEye.

Economic Significance of Improved Environmental Forecasting

Monica Hale, FIEMM FRGS
Sustainability Director
Science Applications International
Corporation (SAIC)
McLean, Va.

Weather conditions and climatic factors affect virtually all economic sectors. Continuous enhancement in accuracy, coverage and resolution of Earth-observing systems by NASA and NOAA will lead to better-informed decisions by end users, thus reducing risk and improving operations.

Some industries are better prepared to utilize enhanced data than others. The energy and financial services sectors—in particular, the insurance industry—use sophisticated forecast models for a range of decision making. The cost of decisions that are based on limited, incomplete or inaccurate data is compounded by errors resulting from deficiencies in the applications or models through which the data are processed. Better understanding of the needs of customers in these industries is imperative.

Use of Environmental Information
in the Energy Sector

Accurate forecasts of power demand are directly reflected in the profitability and commercial competitiveness of energy suppliers and distributors. Better anticipation of power demand—through load forecasting techniques based on weather forecast and market data—is essential to the operations of energy suppliers. Such a capability assists them in matching supply to demand, complying with environmental and other regulatory requirements, and operating in a socially responsible manner.

Load forecasting models assimilate weather forecast data and other information to predict power demand and to mitigate operational risk. However, the accuracy of load forecasting systems is limited by errors in weather data and is compounded by deficiencies in weather and load forecast models and by human error. Reducing load forecasting errors is critical to cost-sensitive areas.

Since deregulation of the energy market, power providers are even more conscious of factors affecting their operations and efficiency. Power generation, transmission and distribution companies increasingly have to deal with issues related to reliability and cost effectiveness. At the same time, they are becoming more and more aware of the overall risks to capital posed by weather conditions and, over the longer term, by climate change.

In fact, weather is often cited as the “tipping point” causing unreliability in the system, decreasing the efficient supply of power and leading to substantial costs. Unpredicted sea breezes, back-door fronts, afternoon thunderstorms and other environmental factors tend to result in overestimated forecast demand. Under-forecasting summer temperature extremes can cause load swings that might cost as much as $75-200 per megawatt hour.(1)

Figure 1 Time scales and operational functions within each timescale.

Heightened power demand is largely driven by high summer temperatures, and to a lesser extent by low winter temperatures. Temperature also acts as a constraint on transmission systems, which are rated to carry specific power loads at defined temperature intervals. Wind, too, has an impact on the system, as it modifies temperature. Hence, weather conditions impact not only overall power usage, but also the carrying capacity of the transmission system.

Figure 2 Map of the U.S. Regional Transmission Organizations (RTO) showing the California (2) and New England (7) Transmission Territories relevant to these case studies.

Energy companies need to know what weather information is available, how to access it and use it to inform operational decision making, and how to employ it for strategic planning. They also need to know which decision support tools are most appropriate and effective. The value of environmental information can be demonstrated by its impact on pricing, scheduling, risk management and, ultimately, the company's bottom line of profitability.

The energy industry utilizes forecast data on different time scales depending on the operational function (see Figure 1).

A series of benchmarking energy sector studies showed these deficiencies:

  1. Error in weather forecast data;
  2. Inadequate incorporation of environmental data into load forecast models;
  3. Load forecast model error;
  4. Misapplication of load forecast output in business transactions.

The economic benefits of environmental observations to the U.S. energy and financial services sectors, and the potential of improved observing systems to increase such benefits, are illustrated primarily by two energy sector case studies.(2) The U.S. energy transmission system is divided into distinct areas, each run by its own independent system operator (see Figure 2.)

California Delta Breeze Predictability

Figure 3 Wind blows from inland westward, funneling through the Golden Gate strait, just north of San Francisco proper. Image taken Aug. 28, 2004 by IKONOS,
courtesy of GeoEye.
During the summer, northwest winds are drawn into California's Central Valley over the Golden Gate strait and into the lower portions of the San Francisco peninsula. To the south of Mount Tamalpais, the northwesterly winds accelerate considerably as they stream through the Golden Gate (see Figure 3). This channeling of the flow through the Golden Gate produces a jet that sweeps eastward but widens downstream, producing southwest winds at Berkeley and northwest winds at San Jose; a branch of this air stream curves eastward through the Carquinez Straits and into the Central Valley. The result of this “delta breeze” is a massive decrease in power demand that can result in a 4,000 megawatt shift in just a few hours. Few energy systems experience such extreme weather phenomena.(3) The delta breeze is strongly influenced by large-scale synoptic weather patterns that move into California from the northern Pacific Ocean.(4)

New England Weather Conditions

A similar study was carried out looking at weather conditions impacting the Independent System Operator for New England (ISO-NE). A case-by-case analysis was completed to identify the standard synoptic weather events that are correlated with high weather forecast errors. In 27 of the 30 cases examined, the load model errors could be explained by weather forecast errors associated with recurring weather features, the key ones being:

  1. A frontal boundary moving through the area slower or faster than expected.
  2. Easterly or northeasterly maritime winds resulting in temperatures below those that were forecast.
  3. Strong westerly winds flowing downhill from the mountains to the coastal plain around Boston, Mass. and Providence, R.I., resulting in compression of the air and, thus, temperatures higher than forecast. (See Figure 4 at top of page.)
  4. Unexpected afternoon thundershowers resulting in temperatures lower than forecast.

A summary of these results is presented in Figure 5.

A review of forecast errors and associated market prices during summer 2002 shows that significant forecast errors occurred during periods of high market prices. This relationship suggests that marketers may be able to pinpoint key load and temperature events, and that there is a need to reduce load forecast errors.

Further investigation indicated that the weather forecast component in the model was responsible for only approximately 40 percent of the overall load error. A subsequent test of the 30 days with the highest load error revealed that when the load error was high, forecast error was responsible for a substantially larger percentage of the load model error. The forecast improvement was only 6.41 percent for the entire model run. This improvement more than doubled to 15.76 percent when the analysis was conducted on the 30 days with the highest error.
Figure 5 Load model errors due to weather forecast errors were found in 27 of 30 cases studied.

These results indicate the potential to improve the load model forecast by improving the weather forecast. It is difficult to estimate the total cost to ISO-NE, or a true social cost, for these errors. However, during some of the highest-error days, including one in which an outage at a major base load plant necessitated expensive imports from Quebec, costs approached $1,000 per megawatt hour.

The weather forecast component and the total load error exhibit minimal correlation (r2 = 0.1900). This means that under-forecasting of the temperature profile does not necessarily imply under-forecasting of the load, and vice versa. The model component and the total load error possessed a high degree of correlation (r2 = 0.8564). This means that the more accurate the temperature forecast, the more likely the power load requirements will be.

Overall, the case studies above suggest that environmental conditions, particularly weather, can have significant impacts on the bottom line of energy supply and delivery organizations. The trend toward real-time pricing and sub-hourly markets makes load forecast accuracy even more critical.

The benefits of improving the day-ahead weather forecast by one degree Fahrenheit include the following:

  1. Approximately $20-25 million per year in cost avoidance for a major Northeast regional transmission authority.
  2. About $1-2 million per year cost avoidance for a large regional distribution company.

Effects on private energy companies' earnings per share (EPS) can be dramatic. One of the large generators in the Northeast reported that its earnings per share in 2003 versus 2002 were positively affected by $+.15/share. In an earlier report, it noted a negative result of $-.05/EPS due to adverse weather effects. Thus, over a short period of time, earnings can swing from a loss of $11 million to a profit of $36 million—a $47 million swing—primarily the result of seasonal weather patterns such as a colder-than-average winter.

Weather forecast error costs suppliers, transporters and consumers. Despite improvements in weather forecasting, load forecast errors still average in the range of 1 to 2 percent per year, and these errors add up to significant annual costs. In addition, extreme weather events often occur during periods when energy costs are normally high.

The Financial Services Sector as Related to the Energy Industry

The financial services sector is closely tied to the energy industry, as it influences the way in which energy companies operate their underwriting investment. The energy sector must respond to pressure from the financial services sector to manage weather-related risk. Improved management would minimize unanticipated costs, reduce high generation costs, meet environmental and supply regulatory requirements, and optimize capital investment, all of which impact profitability.(5)

The key to achieving these objectives is the enhanced use of weather forecast information to transform weather-related risk into probabilistic terms, a concept that is understood by the financial services sector and that facilitates improved management and mitigation of financial risk.

Power companies are continually developing strategic plans to manage and mitigate their diverse risks. While the use of weather derivatives to manage risk has gained considerable momentum in the energy industry over the past five years, there is an alternative mitigation strategy available, namely, to use environmental forecast information more effectively as a supplement to, or replacement of, weather derivatives. This alternative strategy essentially represents a change in paradigm from utilizing environmental information for regulatory compliance to using it for competitive advantage through operational application. Hedging risk with information is the goal.


The case studies indicate that forecast information is now of sufficient accuracy to be incorporated into decision-making processes and models with greater effectiveness. New methods are facilitating the creation of probability trends, which the energy industry will be able to incorporate into its forecast and financial models. The challenge is to shift away from the statistical forecast methods used currently to the probabilistic data. The Northeast case study confirmed the need for this shift, as significant levels of error in both weather forecasting and load forecasting were found.

It is important that the information that is generated from the enhanced observing systems be rapidly assimilated into society to enhance economic competitiveness, improve public safety and address policy imperatives. While it is commendable to produce information that is “capable” of enhancing various industry operations and decision making, it is important that the providers, such as NASA and NOAA, help to prepare users to receive the information and maximize its utility. An investment in understanding the customers' needs is essential to transferring observing systems research to practical operations.End


  1. Mary Altalo and Monica Hale, 2004. “Turning Weather Forecasts into Business Forecasts,” Environmental Finance, May 2004.
  2. Science Applications International Corporation (SAIC) undertook a series of regional and industry-specific studies sponsored by NOAA from 1999 to 2005.
  3. T. D. Davis, D. Gaushell, D. W. Pierce, and M. Altalo, 2005. “Guessing Mother Nature's Next Move: What Can Be Done to Improve Weather Prediction and Load Forecasts?” Public Utilities Fortnightly, August, 2005.
  4. Mary Altalo, Todd D. Davis, and Monica Hale, 2004. “The Economic Benefit of Incorporating Weather and Climate Forecasts into Western Energy Production Management, Final Report,” a project sponsored by the National Oceanographic and Atmos-pheric Administration (NOAA).
  5. Monica Hale, 2003. “Environment, Weather and Climate Information in the Financial Services Sector” In Practice, Bulletin of the Institute of Ecology and Environmental Management (IEEM), No. 42, p. 9-12, Dec. 2003.
Sensors & Systems | Monitoring, Analyzing and Adapting to Global Change | Stay in tune with the transformation. Subscribe to the free weekly newsletter.