Effect of urban warming on blooming of Prunus yedoensis

Abstract An attempt is made to estimate the shift of blooming dates of Prunus yedoensis (the Japanese cherry tree) by urban warming. The ‘number of days transformed to standard temperature’ method is used to estimate blooming dates from temperature data. It is shown that the accuracy of this estimation is good enough to use this method for our purpose. Urban effects on blooming dates are obtained by subtracting the estimated blooming dates calculated using the urban-effect-eliminated temperature from those obtained using the observed temperature. It is found that the amount of shift of blooming date is almost proportional to the amount of urban warming at each site. It is also shown that the proportional constants are closely related to mean blooming dates at each place. The range of the constants is 2.5–5.5 days/°C with smaller values in southern and larger values in northern Japan. General quickening trends due to an increase of urban warming are observed in large cities in Japan. A relatively rapid increase to earlier blooming days was obtained during the 1950s through mid-1970s, then the increasing tendency stopped. In Kyoto, computational results showed a decrease in quickening tendency after the mid-1970s. By comparing variations of actual and estimated blooming dates, it is postulated that this reversal of trend at this site had actually occurred. Methodology and possible sources of errors are discussed. Examples of shifts in number of days of blooming of Prunus yedoensis by urban warming at stations in five large cities in Japan are given.