Golder Daily Climate Record Generator

摘要

Historical meteorological records, while generally being the most accepted inputs for long term hydrology models, often suffer from short periods of record. While these shortages have sometimes been addressed by duplicating records to extend the data period, or using average values, few methods exist to capture the full range of possible meteorological conditions. Synthetic generators (which can be used to address more extreme events) can also be used to extend records; however, these generators tend to require a great deal of climate information which is generally either unavailable or requires advanced subject knowledge. In order to extend climate records in a simple yet meaningful way, Golder has created a synthetic climate generator which requires only statistical information to produce a synthetic climate record. The generator, created using the GoldSim Monte Carlo simulation software, generates daily precipitation and temperature data for a site based on the statistics of the measured climate data. Stochastic processes are used to generate annual, monthly, and daily values that align with the statistical probabilities of the measured data rather than being restricted by the range of measured data. Thus, the generator can be set to generate hundreds of years of data covering a wide range of conditions not reflected in the measured data, yet still fits within the range of likely site conditions. This allows for much more rigorous testing of hydrologic model scenarios and more thorough understanding of the behaviour of engineered systems based on those models.