Bubmann, Sebastian (2021) Rainfall patterns in Port Moresby, Papua New Guinea. (Bachelor thesis), Christian-Albrechts-Universität zu Kiel, Kiel, Germany, 25, III pp.

Text Thesis_Bubmann.pdf - Published Version Restricted to Registered users only Available under License German copyright act UrhG. Download (3MB) | Contact

Abstract

Direct use of climate model precipitation simulations to guide decision making is tempting but bears a significant risk of false conclusions due to potential model biases. This thesis develops a methodology for testing climate model precipitation biases and applies it to the Hadley Centre Global Environment Model version 3 climate model over Port Moresby, Papua New Guinea as a case study. To identify biases, different statistical metrics are compared for spatial and local precipitation observations and historical climate model simulations. Emphasis is given to the assessment of variability, as extreme precipitation events have a documented potential of devastating socio-economic impacts. It is found that the model is on average too dry and supresses extreme rainfall events. To identify possible reasons for the observed reduction in simulated rainfall variability and suppression of extreme rainfall events, a literature study on the Hadley Centre Global Environment Model version 3’s skill to simulate tropical rainfall variability is conducted. The model was reported to simulate the El Nino Southern Oscillation quite well but lacks skill in the simulation of the Madden-Julian Oscillation and its’ effects on convective precipitation. It was concluded that the Madden-Julian Oscillation may be one of the key drivers for the observed precipitation bias. The developed methodology was successfully applied under this study because it revealed a precipitation bias for the case study. The identified bias is significantly impacting the confidence in the models’ simulations of precipitation variability and extreme events over the area of interest. It can be concluded that the simulated data should not be used “out of the box” without any prior correction of the bias. Most likely a bias present in historical simulations will also be an issue for simulations of future climate. Consequently, it is important not to use the raw model climate projection data without a prior bias correction.

Actions (login required)

View Item