Recent studies have highlighted a strong relationship between the spatiotemporal variability of precipitation and large-scale atmospheric circulation, providing valuable insights for effective regional water resource management. However, establishing significant relationships faces notable limitations. The non-stationarity observed in precipitation time series and the intricate behavior of climatic variables pose challenges for accurately capturing correlations using linear approaches. Despite the recognized effectiveness of nonlinear methods in handling oscillatory processes within climatic series, they present difficulties such as high computational costs and challenges in understanding patterns in noisy signals and short records, along with challenges in selecting parameters. In response to these challenges, a new similarity metric based on changepoint locations and trends between changepoints is proposed, with the expectation that it will perform well when the strength of the relationship between predictor and predictand changes with time, as is often observed in teleconnections. The proposed approach can be seen as a flexible non-parametric non-linear relationship between time series. Using the proposed approach, a strong relationship was found between mean seasonal Sea Surface Temperature (SST) anomalies and some climatological stations in California. Results, based on the new similarity measures, notably confirm the influence demonstrated by linear and nonlinear approaches to SST anomalies in the Nino3.4 and Nino 1+2 regions of the equatorial Pacific on precipitation along the west coast of North America, primarily during the winter months. The proposed method is utilized to map regions in the ocean where SST influences specific stations, opening the door to improved seasonal forecasting of extreme precipitation.