CAJ | 학술논문

利用改进的非参数统计检验 Mann-Kendall（MK）方法进行序列趋势检验时，因考虑了时间序列自相关性的影响而使检验结果更趋严格。本文采用MK法及改进的MK法对新疆地区41站1961－2010及2015－2099年主要经济作物棉花和甜菜需水量 ETc和相关气候要素的变化趋势进行了检验，探讨了序列自相关性对 ETc趋势检验结果的影响，并给出了各站及全疆ETc的变化趋势。结果表明：1）当自相关阶数大于0时，应采用改进MK法判断ETc的变化趋势；2）1961－2010棉花和甜菜的ETc具有一定的相依结构，2015－2099年棉花的ETc具有非常明显的长程相关性，但2015－2099年甜菜的ETc均具有时间独立性；3）无论对于棉花还是甜菜ETc，各站点Sen斜率和线性斜率值很接近，满足极好的线性关系，表明两者在反映趋势幅度时具有高度一致性；4）ETc序列是否存在相依性对于趋势显著性有重要影响，最终的趋势检验需用MK法和改进的MK法综合判断。1961－2010和2015－2099年棉花和甜菜的ETc在多数站点呈下降趋势，但其显著性各异。ETc序列的下降说明未来新疆棉花和甜菜用水需求降低，遭受干旱的风险降低。棉花及甜菜ETc的趋势特征对于气候变化背景下干旱-半干旱区作物灌溉制度的制定和防灾抗旱有重要参考价值。
이용개진적비삼수통계검험 Mann-Kendall（MK）방법진행서렬추세검험시，인고필료시간서렬자상관성적영향이사검험결과경추엄격。본문채용MK법급개진적MK법대신강지구41참1961－2010급2015－2099년주요경제작물면화화첨채수수량 ETc화상관기후요소적변화추세진행료검험，탐토료서렬자상관성대 ETc추세검험결과적영향，병급출료각참급전강ETc적변화추세。결과표명：1）당자상관계수대우0시，응채용개진MK법판단ETc적변화추세；2）1961－2010면화화첨채적ETc구유일정적상의결구，2015－2099년면화적ETc구유비상명현적장정상관성，단2015－2099년첨채적ETc균구유시간독립성；3）무론대우면화환시첨채ETc，각참점Sen사솔화선성사솔치흔접근，만족겁호적선성관계，표명량자재반영추세폭도시구유고도일치성；4）ETc서렬시부존재상의성대우추세현저성유중요영향，최종적추세검험수용MK법화개진적MK법종합판단。1961－2010화2015－2099년면화화첨채적ETc재다수참점정하강추세，단기현저성각이。ETc서렬적하강설명미래신강면화화첨채용수수구강저，조수간한적풍험강저。면화급첨채ETc적추세특정대우기후변화배경하간한-반간한구작물관개제도적제정화방재항한유중요삼고개치。
Mann-Kendall (MK) trend test is a nonparametric statistical test. When the modified MK (MMK) test was used for serial trend test, it was considered to be stricter than the MK test considering the influence of self-correlation of time series. In this article, the MK method and the MMK method were combined for trend tests of water requirements (ETc) for two main economic crops, i.e., cotton and sugar beet, at 41 stations in Xinjiang over two periods of 1961-1961 and 2015-2099. The influences of serial self-correlation ofETc on trend test results were investigated, and the spatial distribution of trends inETc of each site and the whole Xinjiang were presented. The results showed that: (1) When the order of self-correlation jwas greater than 0, the statistics obtained from the MMK test should be adopted. (2) Over 1961-2010,ETc of both cotton and sugar beet had certain temporal-dependent structures. Over 2015-2099,ETcof cotton had very significant long-range correlation, butETc of sugar beet hadn’t any temporal dependence. (3) Over 1961-2010, which the trends inETc of cotton at 37 sites decreased, of which the trends inETc of cotton at 8 sites were significant. Over 2015-2099 for A2 scenario, trends inETc of cotton at all of the sites in Xinjiang decreased, of which trends inETc of cotton at 13 sites were significant, while trends inETc at 28 sites were not significant. For B2 scenarios, only at Luntai the trend inETc of cotton increased insignificantly, trends inETc of cotton at the remaining 40 sites decreased. Over 2015-2099 for A2 scenario, trends inETc of sugar beet at 13 sites increased, of which trends at 3 and 10 sites were significant and insignificant, respectively; trends inETc of sugar beet at the remaining 28 sites decreased, of which the trends were significantly and insignificant at half of the 28 sites. For B2 scenario, trends inETc of sugar beet at 9 sites increased. (4) Over 1961-2010, trends inETc of cotton and sugar beet in the whole Xinjiang decreased significant and insignificant, respectively. Over 2015-2099 for A2 and B2 scenarios, both trends inETc of cotton decreased significantly; trends inETc of sugar beet also decreased insignificantly for A2 scenario but significantly for B2 scenario. (5) For both cotton and sugar beet, Sen’s slopeb were related to the linear slopeSL with high coefficients of determination with linear function, which reflected the high consistency of the two variables when describing the trend amplitudes (6) It is necessary to consider the effects of self-correlation forETc series when using MK test. (7) The synthetic effects of decreases in wind speed and solar radiation as well as the increases in precipitation and relative humidity on the restraining ofETc surpassed the effects of increases in air temperatures in Xinjiang. In conclusion, decreases in the trends ofETc series showed that there were reduced water demands of cotton and sugar beet in the future in Xinjiang, and the risk of drought also reduced. Trend variations in ETc for cotton and sugar beet had important reference value for making the crop irrigation planning and drought disaster prevention under the background of climate change in the arid and semi-arid areas.