2nd UF Water Institute Symposium Abstract

Submitter's Name Tirusew Asefa
Session Name Poster Session: Managing Water and Energy in a Transitioning Environment 1
Category Managing water and energy in a transitioning environment
Poster Number 301
Author(s) Tirusew Asefa,  Tampa Bay Water
  Alison Adams,  Tampa Bay Water
  A “Predictor-Correct” Framework for Aggregating Nino3.4 Forecasts for use in Local Rainfall Simulations
  Abstract. Nino 3.4 (120 – 170W and 5N- 5S) Sea Surface Temperature (SST) anomalies are the main indicator of ENSO states. Since January 2005 the International Research Institute (IRI) of Columbia University compiles Nino3.4 forecasts from several Dynamical and Statistical models. These models forecast SST anomalies from three to nine overlapping 3-month periods. Tampa Bay Water uses these forecasts to make seasonal rainfall probability forecasts based on a contingency table that was derived using over 100 years of SST anomalies and rainfall data. Given the forecasted state of ENSO, it is then possible to derive the local seasonal rainfall probabilities. These results are then used to make seasonal water resources management decisions such as stream withdrawals and reservoir operations and maintenance shut downs. However, the expected skills of each of the 22 ENSO model forecasts, based on historical performance, are not the same. At times the variation of the forecasts ranges from La Nina (anomalies less than -0.50C) to El Nino (those above 0.50C) conditions making their practical use limited and uncertain. Here, we propose a “predictor-corrector” framework that aggregates ENSO model forecasts based on recent performances of each model compared to observed values of SST anomalies for use in rainfall simulations rather than the simple arithmetic average that the IRI currently provides. The results will reduce the seasonal forecast uncertainty for rainfall and improve resource use decisions.