ENSO, PNA and NAO Scenarios for extreme storminess, rainfall and temperature variability during the Florida dry season

Hagemeyer (2000) and Hagemeyer and Almeida (2002, 2003, and 2004) developed a dry season climatology (November 01-April 30) of strong extra-tropical (ET) cyclones (or storms) affecting Florida, and a forecast methodology to predict dry season “storminess” or the accumulated passage of ET cyclones in a dry season from the ENSO signal. The authors began publishing their experimental forecasts on the World Wide Web (WWW) in May 2001 for the 2001-02 dry season. Their storminess forecasts have proven quite reliable when the ENSO signal was dominant as during El Nino (above normal storms) or La Nina (below normal storms) conditions. However forecasting storminess during neutral conditions when other teleconnections such as the PNA, NAO, and MJO may play a dominant role remains problematic.

Hagemeyer and Almeida (2005, http://www.srh.noaa.gov/mlb/enso/16th_climate.pdf ) expanded the storminess climatology to include the period from the 1950-51 to the 2002-03 dry seasons (53 seasons), and added dry season average NAO and PNA measures to the regression equations to determine if they improved upon the ENSO-only forecasts. Their results confirmed that the addition of the NAO and PNA teleconnections did slightly improve the dry season storminess forecast on average. However, the most significant improvements occurred in ENSO neutral seasons when the PNA and NAO can have the most influence and entirely ENSO-based forecasts can fail badly. They also found that seasonal rainfall was obviously related to seasonal storminess, but rainfall was more sensitive to the influence of the PNA and NAO teleconnections, except for those cases when ENSO was extreme. In other words, in most seasons PNA and NAO were very relevant to improvement in prediction of rainfall and storminess.

While these results would appear to be good news, they raises even more challenges for users of the forecasts to understand them, access environmental risks, and potential impacts to society. For example, a confident forecast of a strong El Nino or La Nina can be relatively easily explained to, understood by, and acted upon by stakeholders to mitigate impact. However, in the majority of Florida dry seasons ENSO is either weak or neutral and understanding the inter-relationships and prediction of the PNA and NAO is problematic. Indeed, Hagemeyer and Almeida (2005) documented that storminess and rainfall variability can be just as extreme in ENSO neutral dry seasons as during strong El Nino's or La Nina's.

During their research the authors realized that in addition to use in the development of a seasonal forecast much information applicable to education, preparedness, and mitigation could be mined from their storminess/rainfall climatology. To facilitate the understanding and use of the seasonal forecasts there was a definite need to provide more information to assist potential users of the forecasts in understanding forecast uncertainty, the limits of predictability, factors controlling storminess variability, the range of climatic extremes that have been encountered in the past, and a guide to what extreme scenarios might happen in the future. The authors are developing graphics and conceptual models to help users understand the interplay of the major teleconnections that are responsible for much of the variability and extremes in storminess or lack of storminess (i.e. drought) in Florida. Examples of these products will be presented at the conference and posted on our WWW experimental seasonal forecast page.