Statistical Methods for Assessing Changes in Similarity Maps between Locations in Monthly Temperatures and Precipitation over Time
Abstract
The increase in global average air and ocean temperatures, melting of snow and ice, and the rising global average sea level, confirm that warming of the climate is evident. The rate of warming may vary considerably on the regional scale due to topographical features and other systems interacting with atmospheric circulation. Interdependent ecosystems may be affected by a change in the relationship between climates. We develop a method for assessing temporal changes in the relationships among regional climates. This approach is demonstrated with nineteen weather stations from the USHCN located throughout the Rocky Mountains. We explore the potential to detect changes in relationships among aspects of regional climates across different locations from 1913 to 2012. Temperature and precipitation climate indices are used for creating distance matrices between locations for each decade between 1913 and 2012. Multidimensional Scaling is used to 'map' the similarities/dissimilarities among locations and visualize how relationships change across decades. A matrix of pair-wise Pearson correlations between decadal distance matrices is converted to a distance matrix and analyzed using Distance-Based Permutation MANOVA methods for trends over time. The tests suggest strong evidence for systematic changes in the relationships between these locations over time in monthly minimum, maximum, and average temperature while the evidence is weaker for trends in monthly precipitation. This method can be extended to ecological applications, species distribution modeling and other spatiotemporal studies.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFMNG11A1580H
- Keywords:
-
- 3294 MATHEMATICAL GEOPHYSICS Instruments and techniques