新疆地区蒸发皿蒸发量变化及基于小波的周期分析

Change of pan evaporation in Xinjiang and its periods based on wavelet analysis

Φ20 cm和E601型蒸发皿在新疆均有使用,但两种数据序列自观测开始至今均不完整,尤其自2003年以后数据未进行整合和校正,使得对蒸发皿蒸发量数据的使用和深入分析受到限制。本研究基于Φ20 cm(E_(20))和E601型蒸发皿蒸发量(E601)的共同观测期数据,选取新疆地区57个气象站,分析4~10月E_(20)和E601的换算系数K。以数据序列较长的喀什(隶属南疆)和塔城站(北疆)为例,分析了逐日和逐月尺度下K的变化,并将各月K值用于两个典型站2003—2016年期间4月1日~9月30日E_(20)的估算,得出1961—2016年完整的日E_(20)序列。进一步基于复Morlet小波函数对月及年尺度E_(20)的波谱特性和周期变化进行了分析,结果表明:(1)新疆地区E_(20)和E601的换算系数在4~10月期间具有较大的空间差异,南疆K值较北疆大。(2)喀什和塔城站插补后完整的1961—2016年期间日E_(20)序列具有以年为周期的典型变化,月E_(20)在7月最大,年E_(20)均具有明显的增加趋势;日、月及年尺度下喀什站E_(20)均高于塔城站。(3)两站点1~12月E_(20)的主周期和准周期具有2~16 a的波动,年E_(20)的主周期均为7 a,喀什站准周期为3 a和6 a,塔城站准周期为2 a和4a。本研究可为新疆地区蒸发量序列的插补及进一步应用提供参考。

The 20 cm-in-diameter and E601 type evaporation pans have been used in Xinjiang, China, but these two kinds of data sequences so far are incomplete since the start of observation, especially since 2003, the data have not been integrated and corrected ,which restricted the use of pan evaporation data and their further analysis. Based on the co-observation data of 20 era-in-diameter （ E20 ） and E601 pan evaporation （E601 ）, the 57 meteorological stations in Xinjiang were selected and analyzed for the conversion coefficient K between E2o and Erot from April to October. Taking the Kashi （in southern Xinjiang） and Taeheng stations （in northern Xinjiang） （both had relatively long data sequences and are representative） as an example, the changes of K at daily and monthly time scales were analyzed, and used later in the estimations of E20 values for the period from April 1 to September 30 in the span of 14 years （from 2003 to 2016）. Therefore, the complete data sequences of daily E20 for Kashi and Tacheng over the period from 1961 to 2016 were obtained and could be further used in the spectral characteristics and periodic change analysis using the complex Morlet wavelet function. The results showed as follows ： （ 1 ） The conversion coefficient of E20 and E601 in Xinjiang region had significant spatial differences from April to October, and K values were larger in the southern Xinjiang than those in northern Xinjiang. （2） The complete daily E20 data series during the period from 1961 to 2016 at Kashi and Tacheng station had typical periodical changes on yearly basis. The monthly E20 was the largest in July, and the annual E20 had an obvious increase trend. Daily, monthly and annual E20 values in Kashi were all larger than those in Tacheng. （3） The main period and quasi-period of E20 ranged within 2 - 16 years from January to December, and the main period were 7 years for both stations. The quasi-periods at Kashi station were 2 and 4 years, but at Tacheng station they were 3 and 6 years, respectively. This research could provide useful references for the interpolation and further application of pan evaporation in Xinjiang.