UC Berkeley

Evaluation of long-term temporal and spatial climatic changes in mountainous areas is one of the most critical issues in the problem of global changes. A statistical framework to assess the long-term climatic water balance changes has been developed, including the following phases of the data analysis: (1) a time series analysis of daily, monthly, and yearly averaged meteorological parameters (temperature, relative humidity, precipitation, wind speed, etc.), (2) calculations of potential evapotranspiration, aridity index, and actuarial evapotranspiration, as well as the standard precipitation index (SPI) and standard precipitation-evapotranspiration index (SPEI), and (3) climatic zonation of the watershed area based on the temporal trends of ET and SPEI, performed by means of a combination of the hierarchical k-means clustering and Principal Component Analysis. The paper includes the results of the application of the developed statistical framework for 17 locations of meteorological stations at the East River watershed, using the meteorological datasets for the period from 1966 to 2020. Structural time series changes of meteorological variables and calculated water balance parameters are used to determine the time of climatic shifts. Calculations of the potential evapotranspiration are conducted using the Thornthwaite, Hargreaves, and Penman-Monteith equations, and evapotranspiration (ET) is calculated using the Budyko model. The time variations of the meteorological and calculated parameters before and after temporal structural breakpoints are evaluated using parametric and nonparametric statistics. The results of the hierarchical clustering are illustrated using the tree dendrograms and the PCA plots of clusters of the studied sites. A comparison of the results of clustering based on the ET and SPEI data showed no difference between these two types of the zonation. A significant shift in the cluster arrangements for the time periods before and after the temporal structural breakpoints indicate that zonation patterns are driven by dynamic climatic processes, which are variable through time and space. Therefore, the watershed zonation requires periodic re-evaluation based on the structural time series analysis of meteorological data.