For Degree-Day models which can benefit from forecasts beyond the 7-day forecast, we now have several options. These include:
1) 10 year averages (current default option): uses most recent 10-year average PRISM data localized to the currently selected weather station as the extended weather forecast.
2) 30 year averages: uses 30-year Normals (currently 1981-2010) as the extended forecast.
3) Last years data: uses data from last year (so you assume this year will be similar to last year).
4) Data from 2 years ago, uses data from the year before last (so you assume this year will be similar to 2 years ago).
5) NEW extended seasonal (7-month) forecast: uses the NOAA North American Multi-Model Ensemble (NMME) forecast system, which is an ensemble average of seven different extended (7-month) numerical models, all driven in part by scientific understanding of how ocean temperatures affect seasonal climate and weather patterns. Check the N. American NMME temperature and rainfall forecast anomaly maps, and the verification index to see how well their 1-7 month lead time forecasts have performed in the past. Since this forecast system is considered experimental, use with caution and we recommend always comparing this forecast with 10- and 30- year average forecasts. Keep in mind that forecast skill will vary widely depending on region, season, and where we are in the predominate El Nino/La Nina cycle.
6) NEW CFSv2 extended seasonal (3.5-month) forecast: uses the NOAA CFSv2 model to produce rainfall event-driven patterns.