Abstract

We investigated the first and last frost dates using meteorological observation data from Japan. First, we identified long-term trends of retardation (mean: +0.224 day/year) in the first frost date in fall and advancement (−0.228 day/year) in the last frost date in spring using historical frost observation data from 1951 to 2010. Trends determined over 20-year subperiods were distinct from long-term trends and were sensitive to decadal changes in daily minimum air temperature. Second, we proposed a scheme to infer the first and last frost dates from the time series of daily minimum air temperature. After optimization, the first (last) cold date when the daily minimum temperature fell below a temperature criterion of ca. 2 °C yielded the best estimate of the first (last) frost date with an error of 2 days in most cases. However, the overall root mean square errors were 13–16 days because some cases with significant misfits deteriorated the values. Both wind speed and humidity shifted the criterion. Sites with strong winds contributed to a decrease in the temperature criterion. Because this scheme only requires the daily minimum air temperature, it is widely applicable to the quantitative evaluation of the first and last frost dates from given meteorological or climate projection data.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call