利用气候数据确定最佳日最高温度。
Determination of optimum daily maximum temperature using climate data.
机构信息
School of Comprehensive Human Sciences, University of Tsukuba, 1-1-1 Tennaoudai (Building D, #709), 305-8577, Tsukuba, Japan,
出版信息
Environ Health Prev Med. 2007 Sep;12(5):209-16. doi: 10.1265/ehpm.12.209.
OBJECTIVE
The relation between daily maximum temperature and mortality rate has a V-shaped pattern; the mortality rate is lowest at a certain temperature, that is, optimum temperature (OT), and the mortality rate increases when the temperature becomes higher or lower than OT. OT is associated with climate, but the relation between OT and long-term average temperature, which is a frequently used index of climate, had an outlier (Okinawa) even in Japan alone. Our objective is to determine the best climate index for OT estimation.
METHODS
We obtained death certificate data, meteorological data and population data for Japan from relevant government ministries. All the data obtained were from 1972 to 1995 except for Okinawa's mortality data (1973 to 1995). Using smoothing spline with the degree of freedom fixed to 6, we computed the OTs for 47 prefectures in Japan. These OTs were exhaustively compared with percentiles of daily maximum, average, and minimum temperatures, along with the long-term average temperature.
RESULTS
Among the candidates of the best climate index, 80 and 85 percentiles of daily maximum temperatures (Tmax80 and Tmax85) showed the highest correlation coefficient with OT (R>0.9, much higher than the R for the long-term average temperature, i.e., 0.63), and the regression models using Tmax80 and Tmax85 best regressed the OT, that is, the difference between the observed OT and the expected OT was smallest when Tmax80 or Tmax85 was used. Unlike previously used average of daily mean temperature, Tmax80 and Tmax85 made Okinawa a nonoutlier. This characteristic is desirable because Okinawa's being an outlier is due to its maritime climate and the capacity to accommodate a different type of climate may expand the applicability of OT estimation method to wider regions in the world. A direct comparison of OT with Tmax75 to Tmax90 revealed that the difference is smallest for the percentile between Tmax80 and Tmax85.
CONCLUSION
We considered that a daily maximum temperature between Tmax80 and Tmax85 is the best climate index for estimating OT in Japan.
目的
日最高温度与死亡率之间呈 V 形关系；在某个特定温度下，死亡率最低，即最适温度（OT），当温度高于或低于 OT 时，死亡率会增加。OT 与气候有关，但即使在日本，OT 与长期平均温度（一种常用的气候指标）之间的关系也存在异常值（冲绳）。我们的目标是确定估计 OT 的最佳气候指标。
方法
我们从相关政府部门获得了日本的死亡证明数据、气象数据和人口数据。除了冲绳的死亡率数据（1973 年至 1995 年）外，所有获得的数据均来自 1972 年至 1995 年。使用自由度固定为 6 的平滑样条，我们计算了日本 47 个县的 OT。这些 OT 与每日最高、平均和最低温度的百分位数以及长期平均温度进行了详尽的比较。
结果
在最佳气候指标的候选者中，每日最高温度的 80 百分位数（Tmax80）和 85 百分位数（Tmax85）与 OT 的相关系数最高（R>0.9，远高于长期平均温度的 R，即 0.63），并且使用 Tmax80 和 Tmax85 的回归模型最佳地回归了 OT，即当使用 Tmax80 或 Tmax85 时，观察到的 OT 与预期 OT 之间的差异最小。与之前使用的日平均温度平均值不同，Tmax80 和 Tmax85 使冲绳不再是异常值。这一特点是可取的，因为冲绳的异常值是由于其海洋性气候造成的，而容纳不同类型气候的能力可能会扩大 OT 估计方法在世界范围内更广泛地区的适用性。OT 与 Tmax75 至 Tmax90 的直接比较表明，Tmax80 和 Tmax85 之间的百分位数差异最小。
结论
我们认为，Tmax80 至 Tmax85 之间的日最高温度是估计日本 OT 的最佳气候指标。
Zotero 插件