Historical Weather Data Base and Reconstruction of the Climate in Historical Time
Historical Weather Data Base and Reconstruction of the Climate in Historical Time
复制标题
历史天气数据库与历史时期气候重建
DOI：
10.5026/jgeography.102.2_131
复制
发表时间：
1993
影响因子：
3.1
通讯作者：
M. Yoshimura
中科院分区：
文献类型：
--
作者：
M. Yoshimura
为了重建仪器观测前日本的气候演变，建立了历史天气数据库。该数据库的对象是江户时代封建氏族的官方日记或各氏族地方办事处的历史文献中的天气描述。其他的描述是在那些大寺庙、大神社、大农民的日记中，以及在一些私人日记中，考虑到旧日记中对天气现象的描述，它们被分为13个天气条件。在这个数据库中，“天气范围”是指一天的天气条件。一天中天气的大致趋势（HN），或诸如热、冷、干、湿（HT）等显著条件，对于了解每天的天气状况也很重要。伴随现象（TK）和风况（KT， TT， KH）也是要编码的对象。1700年至1889年的每日天气记录主要在收藏中。到目前为止，超过2500年的天气记录已经输入到每日数据库中。该系统的主要文件为PC-DA（古气候资料）和PC-REC（古气候记录）。前者由经过编码的天气数据组成，主要用于检索天气状况、绘制地图或制作表格。后者用片假名（日本汉字之一）描述每天的天气情况。这将支持详细的调查。作为每个月潮湿或干燥状况的指标，“PI”被提出。与降水有一定关系的天气条件可分为三类。每种类型的月频率乘以常数。1代表“小雨”，2代表“雨”，5代表“大雨”。每个月每一天的总值为“PI”，便于在同一日序列内进行比较。各季九州小仓市PI与无印良市降水散点图如图4所示。6个城市的PI时间序列显示了日本夏季降水存在不同的行军（图5）。图6重建了1781 - 1789年夏季月份的月况。
In order to reconstruct the march of climate before instrumental observations in Japan, Historical Weather Data Base has been constructed. Weather descriptions in historical documents, such as official diaries of Feudal Clans in Edo Era, or of local offices in each clan, are objects of the data base. Other descriptions are in those diaries of large temples, large shrines and large farmers, and in some of private diaries Considering the description of weather phenomena in old diaries, they were grouped into 13 weather conditions In this data base, “weather range” means weather conditions in a day. A rough tendency of the weather in a day (HN), or remarkable conditions such as hot, cold, dry and wet (HT), are also important to understand weather conditions for each day. Accompanied phenomena (TK) and wind conditions (KT, TT, KH) are also objects to encode. Daily weather records for the years from 1700 to 1889 are mainly under collections. So far, more than 2, 500 years weather records have been inputted in the daily data base. The main files of the system are PC-DA (paleo climatic data), and PC-REC (paleo climatic record). The former consists of encoded weather data which will be used chiefly to retrieve weather conditions and to draw maps or to make tables. The latter contains weather descriptions in each day with Katakana, one of the Japanese characters. This will support detailed investigationAs an index of wet or dry condition for each month, “PI” is proposed. Weather conditions having some relations with precipitation are grouped to three kinds. Monthly frequencies of each type are multiplied by constants. These are 1 for “little rain”, 2 for “rain” and 5 for “much rain”. The total value in each month in each day is “PI”, which is useful for the comparison within the same daily series. The scatter diagrams of PI in Kokura Kita Kyushu City and Precipitation at Moji for each season are shown in Fig. 4. Time series of PI in 6 cities show the existence of different march in summer precipitations over Japan (Fig. 5). The monthly conditions in summer months from 1781 to 1789 are reconstructed in Fig. 6.