A method for predicting the uncompleted climate transition process

Abstract

Climate change is expressed as a climate system transiting from the initial state to a new state in a short time. The period between the initial state and the new state is defined as the transition process, which is the key part for connecting the two states. By using a piece-wise function, the transition process is stated approximately (Mudelsee, 2000). However, the dynamic processes are not included in the piece-wise function. Thus, we proposed a method (Yan et al., 2015, 2016) to fit the transition process by using a continuous function. In this paper, this method is further developed for predicting the uncompleted transition process based on the dynamic characteristics of the continuous function. We introduce this prediction method in detail and apply it to three ideal time sequences and the Pacific Decadal Oscillation (PDO). The PDO is a long-lasting El Niño-like pattern of Pacific climate variability (Barnett et al., 1999; Newman et al., 2016). A new quantitative relationship during the transition process has been revealed, and it explores a nonlinear relationship between the linear trend and the amplitude (difference) between the initial state and the end state. As the transition process begins, the initial state and the linear trend are estimated. Then, according to the relationship, the end state and end moment of the uncompleted transition process are predicted.