Dynamic Time Warping as a Means of Assessing Solar Wind Time Series

Abstract Over the last decades, international attempts have been made to develop realistic space weather prediction tools aiming forecast conditions on Sun and in interplanetary environment. These efforts led development of appropriate metrics assess performance those tools. Metrics are necessary validate models, compare different monitor improvements a certain model over time. In this work, we introduce dynamic time warping (DTW) as an alternative way evaluating models and, particular, quantifying differences between observed modeled solar wind series. We present advantages drawbacks method, well its application Wind observations EUHFORIA predictions at Earth. show that DTW can warp sequences time, align them with minimum cost by using programming. It be applied for evaluation series two ways. The first calculates sequence similarity factor , number provides quantification how good is compared ideal nonideal scenario. second quantifies amplitude points best matched sequences. As result, serve hybrid metric continuous measurements (e.g., correlation coefficient) point-by-point comparisons. promising technique assessment profiles, providing once most complete portrait model.