Detecting and Adjusting Temporal Inhomogeneity in Chinese Mean Surface Air Temperature Data

Adopting the Easterling-Peterson (EP) techniques and considering the reality of Chinese meteorological observations, this paper designed several tests and tested for inhomogeneities in all Chinese historical surface air temperature series from 1951 to 2001. The result shows that the time series have been widely impacted by inhomogeneities resulting from the relocation of stations and changes in local environment such as urbanization or some other factors. Among these factors, station relocations caused the largest magnitude of abrupt changes in the time series, and other factors also resulted in inhomogeneities to some extent. According to the amplitude of change of the difference series and the monthly distribution features of surface air temperatures, discontinuities identified by applying both the E-P technique and supported by China's station history records, or by comparison with other approaches, have been adjusted. Based on the above processing, the most significant temporal inhomogeneities were eliminated, and China's most homogeneous surface air temperature series has thus been created. Results show that the inhomogeneity testing captured well the most important change of the stations, and the adjusted dataset is more reliable than ever. This suggests that the adjusted temperature dataset has great value of decreasing the uncertaities in the study of observed climate change in China.

Adjusted Log-rank Test with Double Inverse Weighting under Dependent Censoring

It is a common issue to compare treatment-specific survival and the weighted log-rank test is the most popular method for group comparison. However, in observational studies, treatments and censoring times are usually not independent, which invalidates the weighted log-rank tests. In this paper, we propose adjusted weighted log-rank tests in the presence of non-random treatment assignment and dependent censoring. A double-inverse weighted technique is developed to adjust the weighted log-rank tests. Specifically, inverse probabilities of treatment and censoring weighting are involved to balance the baseline treatment assignment and to overcome dependent censoring, respectively. We derive the asymptotic distribution of the proposed adjusted tests under the null hypothesis, and propose a method to obtain the critical values. Simulation studies show that the adjusted log-rank tests have correct sizes whereas the traditional weighted log-rank tests may fail in the presence of non-random treatment assignment and dependent censoring. An application to oropharyngeal carcinoma data from the Radiation Therapy Oncology Group is provided for illustration.