Advisor

Heejun Chang

Date of Award

Summer 9-7-2017

Document Type

Thesis

Degree Name

Master of Science (M.S.) in Geography

Department

Geography

Physical Description

1 online resource (vi, 51 pages)

Subjects

Rainfall intensity duration frequencies -- Oregon -- Portland -- Case studies, Climatic changes -- Risk assessment, Rainfall frequencies

DOI

10.15760/etd.5777

Abstract

In response to increased greenhouse gases and global temperatures, changes to the hydrologic cycle are projected to occur and new precipitation characteristics are expected to emerge. The study of these characteristics is facilitated by common indices to measure precipitation and temperature developed by the Expert Team on Climate Change Detection and Indices (ETCCDI). These indices can be used to describe the likely consequences of climate change such as increased daily precipitation intensity (SDII) and heavier rainfall events (R95p). This study calculates a subset of these indices from observed and modelled precipitation data in Portland, Oregon. Five rainfall gages from a high resolution rain gage network and projections from three downscaled global climate models including CanESM2, CESM1, CNRM-CM5 are used to calculate precipitation indices. Mann-Kendall's tau is used to detect monotonic trends in indices. The observational record is compared with models for the historic period (1977-2005) and these past trends are compared with projected future trends (2006-2100). The influence of study unit on trend detection is analyzed by computing trends at the annual and monthly scale. Study unit is shown to be important for trend detection. When the annual study unit is used, projected future trends towards increased precipitation intensity and event volumes are not observed in the historic data. However, when analyzed with a monthly study unit, trends towards increased precipitation intensity and event volumes are observed in the historic data. These trends are shown to be important for Portland area flooding, as precipitation indices are shown to significantly correlate with 40 maximum peak flow events that occurred during the period of study.

Persistent Identifier

http://archives.pdx.edu/ds/psu/21759

Share

COinS