Publication Date



Open access

Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PHD)


Meteorology and Physical Oceanography (Marine)

Date of Defense


First Committee Member

Ben P. Kirtman

Second Committee Member

Amy C. Clement

Third Committee Member

Brian J. Soden

Fourth Committee Member

Randal D. Koster

Fifth Committee Member

Siegfried D. Schubert


Monthly and seasonal climate prediction of variables such as precipitation, temperature, and sea surface temperature (SST) is of current interest in the scientific research community, and also has implications for users in the agricultural and water management domains, among others. This dissertation studies a variety of approaches to seasonal climate prediction of variables over North America, including both climate prediction systems and methods of analysis. We utilize the North American Multi-Model Ensemble (NMME) System for Intra-Seasonal to Inter-Annual Prediction (ISI) to study seasonal climate prediction skill over North America. We also use the Community Climate System Model version 4.0 (CCSM4) to preformed targeted climate prediction experiments to study contributions to skill or predictability from SSTs, land and atmosphere initialization, and ocean-atmosphere coupling. While all can be considered important for predictions, we show that for winter predictions, SST errors are a leading cause in forecast degradation, and improvement of SSTs causes a significant improvement in skill. Climate models, including those involved in NMME, typically overestimate eastern Pacific warming during central Pacific El Niño events, which can affect precipitation predictions regions that are influenced by teleconnections, such as the southeast US. Land and atmosphere initialization, and the minimization of errors in these initial states, shows moderate improvement in skill, expected for the first seasonal lead. Finally, ocean-atmosphere coupling, in the context of this experiment design and in relation to prescribed SST versus fully coupled hindcasts, is a comparatively weak contribution to prediction skill and predictability.


climate; climate prediction; predictability; teleconnection; precipitation; temperature