Trends in extreme precipitation indices across China detected using quantile regression.

For China, long-term changes are detected not only in the means of eight extreme precipitation indices, but also in their distribution shapes by quantile regression. This resulted in different trends for the means and other aspects of the index distributions. The differences between changes in the means and upper/lower extremes vary with region and index. A noteworthy feature is that changes in upper tails of the index distributions across a broad area, especially in the south, are at a much higher rate than mean trends estimated by the traditional linear regression model. This has practical implications for disaster risk management. [ABSTRACT FROM AUTHOR]