Prescient Weather, the creator of CropProphet, also offers a long-range forecast and analysis tool called the World Climate Service. Our company focuses on developing market-leading, scientifically-based climate information services. CropProphet is an excellent example of a climate information application because annual crop yields are dependent on weather conditions over an entire crop season. In other words, crop yields are based on the climate of a crop season. The mission of CropProphet is to quantify the impact that crop season weather conditions have on crop yield and production.
The World Climate Service focuses on long-lead forecasts in the subseasonal (i.e. several weeks ahead) and seasonal (a few months ahead) time frames. This post highlights forecast information from the World Climate Service that demonstrates that the U.S. precipitation extremes of the second half of May 2019 were, in fact, predictable.
The analysis below shows the total precipitation across the United States during the 14 days ending June 3, 2019, expressed as a percent of normal.
An analysis of accumulated precipitation is often most usefully presented in terms relative to climatology, or long-term normals, as shown below on the left side. For the month of May as a whole, this map shows that some regions of northern Oklahoma, southern Kansas, and northern Missouri received as much as 3 times the normal quantity of precipitation. The map on the right shows a corn yield analysis from CropProphet, illustrating that these regions are currently experiencing a significant impact in terms of departure from normal yields.
The markets have reacted very strongly to the likely impact of the extreme May rainfall on final US corn yield and production. Imagine the value of knowing that this precipitation event was going to happen weeks in advance.
Below is an analysis of the ECMWF (European Centre for Medium Range Weather Forecasting) extended ensemble forecast issued on May 13, 2019, for the period May 21, 2019, to June 3, 2019.
Compare the forecast to the first image in the post. The forecast was issued 7 days in advance of the beginning of the target period and 19 days before the end of the target period.
The World Climate Service does not simply provide raw model forecast output. It’s been carefully tuned, optimized, and calibrated to extract as much predictability as possible from the information available. As demonstrated by this example, long-lead weather and climate science has been evolving rapidly and is now at the point that modern long-range forecasts have tremendous value for sophisticated users of the information.