Inter-Annual Distributions of Rainfall in Hawai’i: Applying a Geospatial Approach on a Single Source Rain Shadow

Hawai’i (The Big Island) has one of the most intricate and compact climate systems in the world, owing to drastic elevation differences ranging from sea level to 4,207m at the summit of Mauna Kea and with the island being embedded in a relatively consistent easterly trade wind pattern. This results in Hawai’i hosting several climatic zones including the tropical rainforest, tropical monsoon, tropical savannah, oceanic and subpolar oceanic, Mediterranean, arid and semi arid, and tundra climates in an area of 10,000 square kilometers. Given the small size of the island, synoptic scale meteorology can be assumed to be consistent across the entire island and held constant. Moreover, the island is a shield volcano of a conical shape, drastically reducing the complexity of the more complicated rain shadows generated by mountain ranges. These two factors of small size and comparatively simple geometry makes Hawai’i an excellent case to study the climatic effects of both external forcing of precipitation, moisture and wind, as well as variation in the rain shadow effect related to varying conditions over climatological timescales. This study utilized geospatial analysis and divided the island of Hawai’i into four ordinal quadrants and correlated anomalies in rainfall over a 21 year time series, finding differing levels of correlation between each quadrant pair, ranging from 0.74 at the strongest and 0.37 at the weakest. Furthermore, the study found multiple unique patterns in rainfall anomalies when plotted using a polar coordinate system consisting of meso-network stations traveling clockwise around the island. The usage of IDW interpolation solely utilizing the available stations on the island was found to be ineffectual due to poor spatial resolution and the complexity of rainfall patterns. The key challenges in this study was the large amount of data points which had to be removed during the quality control process due to missing measurements, a uniquely challenging involvement of microscale phenomena and reanalysis data not having the resolution needed to be practical in supplementary interpolation. The most successful reanalysis study to date is available by the University of Hawai’i, and demonstrated modest correlations with the meso-network stations used within this study between reanalysis and observed precipitation. The data however still exhibited significant smoothing of extremes, further highlighting the complexity of accounting for highly localized events in the analysis process. Further studies are being developed to better understand the aspects and behavior of climate systems such as the Hawai’i.