Investigating a novel feature of multi-phase rotated empirical orthogonal function to capture spatiotemporal temperature variations

Climate change has profoundly impacted the Northwest Himalayan (NWH) region, necessitating the analysis of its swiftly evolving spatiotemporal dynamics using appropriate statistical approaches. To represent systematic patterns and explain the localized characteristics of the region, we employ Rotated Empirical Orthogonal Functions (REOFs) on mean temperature data (1981-2021). The NWH region is divided into the snow and non-snow cover regions via REOF analysis, allowing for a more accurate representation of the data. In particular, the western parts of Uttarakhand (U.K.) and Himachal Pradesh (H.P.) showed significant fluctuations in the second REOF mode of mean temperature, which accounted for 38.7% of the entire variance. But the principal component (PC) that conveys information about the patterns in these regions fails to accurately reflect the original patterns present in the data. To achieve this, we have introduced Multi-phase REOF analysis for the first time to identify PCs that accurately represent the original data and provide insight into the specific temporal variations. The superiority of Multi-phase REOF analysis over classical REOF analysis is demonstrated through the application of the Mann-Kendall test. To illustrate that the method is not reliant on specific datasets, we have utilized data sets from both NASA and CRU, covering various time periods. Additionally, potential reasons for the observed variances in these specific NWH locations are investigated. Leveraging the derived physical properties and identified temperature zones from the data can enhance climate change modeling efforts in the region.