Long Island Sound temperature variability and its associations with the ridge–trough dipole and tropical modes of sea surface temperature variability

Abstract

Abstract. Possible mechanisms behind the longevity of intense Long Island Sound (LIS) water temperature events are examined using an event-based approach. By decomposing an LIS surface water temperature time series into negative and positive events, it is revealed that the most intense LIS water temperature event in the 1979–2013 period occurred around 2012, coinciding with the 2012 ocean heat wave across the Mid-Atlantic Bight. The LIS events are related to a ridge–trough dipole pattern whose strength and evolution can be determined using a dipole index. The dipole index was shown to be strongly correlated with LIS water temperature anomalies, explaining close to 64 % of cool-season LIS water temperature variability. Consistently, a major dipole pattern event coincided with the intense 2012 LIS warm event. A composite analysis revealed that long-lived intense LIS water temperature events are associated with tropical sea surface temperature (SST) patterns. The onset and mature phases of LIS cold events were shown to coincide with central Pacific El Niño events, whereas the termination of LIS cold events was shown to possibly coincide with canonical El Niño events or El Niño events that are a mixture of eastern and central Pacific El Niño flavors. The mature phase of LIS warm events was shown to be associated with negative SST anomalies across the central equatorial Pacific, though the results were not found to be robust. The dipole pattern was also shown to be related to tropical SST patterns, and fluctuations in central Pacific SST anomalies were shown to evolve coherently with the dipole pattern and the strongly related East Pacific–North Pacific pattern on decadal timescales. The results from this study have important implications for seasonal and decadal prediction of the LIS thermal system.

Continuum‐based teleconnection indices of United States wintertime temperature variability

“…The VPS is least correlated to the AO‐filtered dipole index, which is not surprising because this index was derived to be uncorrelated with the AO index whose DJF relationship with the VPS is moderate ( r = .58). The LIS dipole index is negatively related with the VPS during the late and early winter (e.g., NDJ), reflecting the relationship between the AO and LIS dipole index during that season (Schulte and Lee, 2019). The cross‐correlation plots also indicate that there is lagged relationship between the winter VPS and springtime dipole indices but understanding this connection is beyond the scope of this paper.…”

“…The LIS dipole index is defined as the 500‐hPa geopotential height anomaly at 42.5° N and 75° W minus the 500‐hPa geopotential height anomaly at 70° N and 157.5° W so that positive phases are associated with ridging across the eastern US and troughing over Alaska (Figure 2). One advantage of using this LIS dipole index is that it is defined in all months and is strongly correlated with the EP‐NP index during the cool season (Schulte and Lee, 2019). Thus, this EP‐NP‐like pattern approximates the December EP‐NP pattern, which was suggested to exist in December by Schulte and Lee (2019) even though it cannot be extracted from an EOF analysis.…”

“…One advantage of using this LIS dipole index is that it is defined in all months and is strongly correlated with the EP‐NP index during the cool season (Schulte and Lee, 2019). Thus, this EP‐NP‐like pattern approximates the December EP‐NP pattern, which was suggested to exist in December by Schulte and Lee (2019) even though it cannot be extracted from an EOF analysis.…”

“…(2017) used the continuum approach to construct an eastern North American SLP dipole index that can explain more precipitation variability than the well‐known PNA and NAO indices. In another study, Schulte and Lee (2019) created a ridge‐trough dipole index that explains up to 80% of wintertime temperature variability across the eastern US, rendering it a better predictor of temperature than the well‐studied AO pattern whose relationship with wintertime temperature across the northeastern US is weak. Although those studies demonstrated the ability of the continuum approach to extract relevant teleconnection patterns, the studies focused on a specific region of the US so that those patterns are not optimized to explain temperature variability across other US regions.…”

A Waveform Skewness Index for Measuring Time Series Nonlinearity and its Applications to the ENSO-Indian Monsoon Relationship

“…The waveform skewness index was constructed as follows. First standardized anomaly time series were decomposed into positive and negative events using an event decomposition approach (Schulte and Lee, 2019), where positive (negative) events are contiguous strings of positive (negative) anomalies. The peak intensity of a positive (negative) event was defined as the maximum (minimum) value 20 obtained by the data points associated with the event.…”