Description
In the first part of this thesis, anomalies and errors are computed for four different datasets from the United States Historical Climatology Network Version 2 (USHCN V2) for the contiguous United States. The used datasets are mean temperature, maximum temperature, minimum temperature and precipitation. The error variance calculations are an important part of climate tracking and are performed for gridded mean temperature anomalies from the USHCN V2 data for each of the four datasets. The error computations are performed for each grid box in a 3.5° longitude by 2.5° latitude grid for the USHCN V2 station network and for each month from January 1895 to December 2009. Error variances are regionally and then annually averaged to create one error value that represents the contiguous United States for a particular year. In the second part of this thesis, an interface was developed that automatically performs all the computations needed in order to update the results of the first part of this thesis. Furthermore, anomaly and error variance data computed in the first part of this thesis are used to make inferences about climate change. Nonparametric statistical methods are employed to create rankings for the hottest and coldest years as well as years with the most and least precipitation. Furthermore trends for all the four datasets, i.e. temperature and precipitation are estimated and then tested with nonparametric methods. Finally, the results are deposited at the US National Climatic Data Center (NCDC)
