Assesment Of The Long-Term Changes In Sub-Daily Precipitation In A Tropical Complex-Terrain Region: Extreme Events ​​And Their Relation With Temperature Increase

During the last few years, the debate over the changes in the global average temperature has been one of the most important political issues. A better understanding of the effects of the temperature increase on other variables is one of the main challenges in climate. In the last IPCC report, due in part to the limited rainfall data in the tropics, there is no reliable conclusion about the observed and expected long-term precipitation changes in association with the global temperature increase. Since water resources are essential for energy generation and food production in the Department of Antioquia, Colombia, a better understanding of changes in rainfall in the long term becomes a vital tool for decision making. This research presents an assessment of rain variability at different temporal and spatial scales over Colombia, and more specifically over Antioquia. The data used was corresponds to long-term records of 86 rain gauges, in addition to 9 temperature stations, and TRMM precipitation products. The use of in-situ rain gauge information allows focusing on a spatial scale useful not only for a general understanding of precipitation changes but also for engineering and other practical applications. Analyses reveal that while there are no long-term trends in precipitation at the monthly or longer timescales, relatively short-lived extreme events show long-term changes in intensity and frequency. Results show that the shorter the duration of the intense events, the higher the magnitude of the increasing intensity trend. Similarly, for more intense events, the trends are also larger and more significant from a statistical point of view. Analysis of temperatures shows a clear relationship with extreme precipitation events with scaling features explained via the Clausius-Clapeyron relation, controlling the intensification of precipitation. The long-term rainfall trends are compared with modeling results from the different scenarios of a small set of CMIP runs given that most models do not adequately represent Colombia´s precipitation climatology. The results indicate a substantial reduction of return period of extreme events with implications in engineering: the current hydraulic designs would be obsolete in less than 50 years if the increment in the frequency of intense events is not considered in the design.