Value of Forecasts

Photo Source: People's Daily

I investigate the social impacts of weather and climate forecasts, one of the most common public goods in the modern world, which has gained increasing significance in the context of climate change. Interacting with scientific development, policy choices, and changes of public expectations, weather and climate forecasts are important in displaying and estimating the value of science and technology advancements, public information improvement, and increased capacity of climate adaptation. My work emphasizes empirical evidence and quantitative analysis with economic models, to outline the significance of accurate weather forecasts in assisting people better adapt to near-future weather shocks. This provides valuable insights for scientists and policymakers on optimizing forecast improvements for social benefit under climate change.

My published and ongoing research focuses primarily on the value of short-term weather forecasts in China, which are produced by an uniform authority source and widely recognized among the large population. I collect a novel dataset of day-ahead, city-level daily weather forecasts reported by the country's top viewership weather program on TV through video transcription with Google Speech-to-Text API. In my paper "The Value of Weather Forecasts: Evidence from Labor Responses to Accurate Versus Inaccurate Temperature Forecasts in China", I demonstrate that labor respond to uncomfortable temperatures (e.g., extreme heat) by reducing hours worked per day given accurate 24-hour weather forecast information, but do not respond given inaccurate forecasts. Based on the revealed preference evidence, I construct a valuation model that incorporates labor decision-making based on forecast and forecast quality to estimate that accurate weather forecasts in China are worth a substantial annual monetary value from the labor sector alone, comparable to the national meteorological service's budget. My other ongoing projects use the same forecast dataset to analyze the effects of accurate forecasts in different sectors. In particular, I have found that accurate cold temperature forecasts reduce negative sentiments, as measured by a sentiment index derived from NLP (natural language processing) analysis of posts on the country's largest social network platform (Weibo). Accurate cold weather forecasts also produce social values in the transportation sector. For instance, road congestion and private car accident claims are reduced under accurate cold forecasts, especially during rain.

In addition, I study the impacts of forecasts for extreme events, particularly those exacerbated by climate change, in both the U.S. and China. Driven by increasing smoke-related intense air pollution due to forest fires in North America, my working paper ``Forecasts and Adaptation to Smokes: Evidence from Residential Electricity and Water Consumption under Smoke Exposure and Smoke Forecasts in California'' studies how people adapt to smoke events given air quality forecasts. I find that accurate 24-hour NOAA smoke forecasts and local air district alerts lead to changes in electricity and water usage during smoke events. For example, under light-to-medium smoke conditions or COVID stay-at-home orders, residential electricity consumption increases for air purification, and water consumption decreases due to polluting water concerns. Under heavy smokes, electricity consumption drops, reflecting more evacuations, and water consumption rises, implying house cleaning needs. I am also working on a project examining the effects of typhoon forecasts in East and Southeast Asia, particularly China. This research combines historical typhoon route forecasts, local precipitation forecasts, and nightlight data to identify the magnitude and duration of electricity usage reduction when typhoons sweep through the region from summer to fall, and to estimate the economic costs of typhoons that could be mitigated by advanced forecasting. The findings will offer policy implications for investment in more accurate typhoon forecasts, improved cross-country meteorological knowledge sharing, and better electricity infrastructure.