Episodic interannual climate oscillations and their influence on seasonal rainfall in the Everglades National Park

The restoration of the Everglades in Florida is an exemplary ecosystem project. A basic challenge of the restoration project is to operate the hydrologic control structures in a manner that allows the right quantity and quality of water to be delivered at the right times to the right locations. An understanding of long-term variations in seasonal rainfall as well as prospects for the upcoming season are of interest for operational planning. This paper aims to characterize the interannual variability in seasonal rainfall in the Everglades and to identify regions of Pacific and Atlantic oceans whose sea surface temperatures (SSTs) may be the carriers of the low-frequency information associated with Everglades rainfall. It is now known that interannual and interdecadal quasi-oscillatory phenomena modulate continental rainfall in many places. The amplitudes of these "oscillations" vary with time, and they conform to activity in specific frequency bands. The dominant low-frequency modes also vary by season. Identifying the climate modes that influence specific low-frequency aspects of rainfall is a challenge that is addressed here using wavelet analysis to diagnose the time-varying low-frequency structure and independent component analysis to identify the spatial modes of variation of the low-frequency signals. The combined approach is termed wavelet-independent component analysis (WICA). In addition to identifying dominant timescales of quasi-oscillatory phenomena that modulate interannual rainfall in the Everglades National Park, we investigate how the amplitude (power) associated with these interannual modes varies at decadal or longer timescales. The analyses presented motivate the need for the development of methods for the analysis and simulation of nonstationary hydroclimatic phenomena. The connection between the resulting low-frequency rainfall modes and sea surface temperatures (SSTs) is then established using correlation analysis using concurrent and preceding season SSTs. The results provide the motivation for the development of a new generation of simulation and forecasting models for rainfall that could directly use such low-frequency information.