黑河流域非一致性极端高温频率特征分析
作者简介:王月华(1992- ),女,河北沧州人,硕士,从事水文学及水资源研究。E-mail:2105140007@cugb.edu.cn
收稿日期: 2016-10-20
要求修回日期: 2017-01-20
网络出版日期: 2017-04-20
基金资助
中央高校基本科研业务费专项资金项目(35832015028)
北京高等学校青年英才计划项目(YETP0654)
Non-stationary frequency analysis of extreme hightemperature in the Heihe River Basin
Received date: 2016-10-20
Request revised date: 2017-01-20
Online published: 2017-04-20
Copyright
王月华 , 李占玲 , 赵韦 . 黑河流域非一致性极端高温频率特征分析[J]. 地理研究, 2017 , 36(4) : 755 -764 . DOI: 10.11821/dlyj201704013
Studying the characteristics of extreme high temperature events under the changing environment is important for the mitigation and adaptation of climate change, as it provides theoretical basis for local disaster prevention and mitigation. How to quantify the non-stationary extreme high temperature and its changes has not well established so far. In this paper, two extreme high temperature indices, i.e. annual mean maximum temperature (AMMaxT) and high temperature intensity (HI), are proposed to describe the extreme high temperature events in the Heihe River Basin. Daily temperature observations from 1960 to 2010 of nine meteorological stations in the Heihe River Basin are collected. Four different statistical tests methods (including Mann-Kendall test, Spearman rank correlation test, Rank sum test and Pettitt test) are employed to detect the non-stationary characteristics of the two extreme high temperature indices. Eight theoretical probability distribution models (including Bate, Gamma, GEV, GPD, Log-Logistic, Lognormal, Wakeby and Weibull) are used to fit the frequency characteristics of the two indices. Trend analysis and change point detection show that nearly all the nine stations have experienced significant trends and obvious change points in both AMMaxT and HI series, and the main variation type is change point variation. Since the theoretical probability distribution models are commonly used to fit the stationary series, the non-stationary AMMaxT and HI series in this study are modified to be stationary by means of the backward restore for consistency. All of the eight probability distribution models can give good fittings to the modified AMMaxT series, while only three of the eight models, i.e. the GEV, GPD and Wakeby models give satisfactory fittings to the modified HI series. According to the ranking of goodness of fit, the GEV and Wakeby models perform the best for both AMMaxT and HI series. Considering its wide applications in other related researches, the GEV model is finally selected as the optimum theoretical one for fitting the extreme high temperature indices in the study area. Based on the GEV model, we calculate the estimated return levels for both the modified and non-modified extreme series at different return periods, and assess the changes of the extreme series at three different return periods (i.e. 10-year, 20-year and 50-year). Overall, the estimated return levels for non-modified extreme series are greater than those for the modified series. This means that the extreme high temperature indices in the study area present trends of increased intensity, shortened return period and increased frequency, which is consistent with the changes of temperature in Northwest China.
Fig. 1 Locations of meteorological stations in the Heihe River Basin图1 黑河流域各气象站点分布图 |
Tab. 1 Basic information of meteorology stations in the Heihe River Basin表1 黑河流域各气象站点基本信息 |
| 站点 | 纬度 | 经度 | 海拔高度(m) | 多年平均气温(℃) | 多年平均日最高气温(℃) | |
|---|---|---|---|---|---|---|
| 上游 | 托勒 | 38°48′ | 98°25′ | 3367.0 | -2.5 | 6.7 |
| 野牛沟 | 38°25′ | 99°35′ | 3320.0 | -2.8 | 7.0 | |
| 祁连 | 38°11′ | 100°15′ | 2787.4 | 1.2 | 10.3 | |
| 中游 | 山丹 | 38°48′ | 101°05′ | 1764.6 | 6.6 | 14.8 |
| 张掖 | 38°56′ | 100°26′ | 1482.7 | 7.5 | 15.9 | |
| 酒泉 | 39°46′ | 98°29′ | 1477.2 | 7.6 | 15.0 | |
| 高台 | 39°22′ | 99°50′ | 1332.2 | 7.9 | 16.1 | |
| 下游 | 鼎新 | 40°18′ | 99°31′ | 1177.4 | 8.5 | 16.6 |
| 额济纳旗 | 41°57′ | 101°04′ | 940.5 | 9.0 | 16.8 |
Tab. 2 Results of trend and change point tests for extreme high temperature indices表2 极端高温指数趋势变异、跳跃变异检验结果 |
| 站点 | 趋势变异 | 跳跃变异 | ||
|---|---|---|---|---|
| AMMaxT | HI | AMMaxT | HI | |
| 托勒 | 3 | 3 | 1989 | 1993 |
| 野牛沟 | 3 | 3 | 1991 | 1993 |
| 祁连 | 3 | -3 | 1996 | 1996 |
| 山丹 | 3 | 3 | 1989 | 1993 |
| 张掖 | 3 | 3 | 1989 | 1993 |
| 酒泉 | 3 | 3 | 1996 | 1993 |
| 高台 | 3 | 3 | 1996 | 1995 |
| 鼎新 | 3 | 3 | 1989 | 1993 |
| 额济纳旗 | 3 | 3 | 1988 | 1996 |
Tab. 3 Coefficients of efficiency for trend and change point tests of extreme high temperature indices表3 极端高温指数趋势变异、跳跃变异效率系数 |
| 站点 | AMMaxT | HI | ||
|---|---|---|---|---|
| 趋势 | 变点 | 趋势 | 变点 | |
| 托勒 | 0.448 | 0.441 | 0.146 | 0.226 |
| 野牛沟 | 0.182 | 0.362 | 0.174 | 0.259 |
| 祁连 | 0.119 | 0.440 | - | 0.192 |
| 山丹 | 0.331 | 0.447 | 0.058 | 0.173 |
| 张掖 | 0.381 | 0.476 | 0.221 | 0.358 |
| 酒泉 | 0.267 | 0.420 | 0.181 | 0.486 |
| 高台 | 0.282 | 0.472 | 0.047 | 0.257 |
| 鼎新 | 0.327 | 0.623 | 0.122 | 0.250 |
| 额济纳旗 | 0.242 | 0.324 | 0.115 | 0.298 |
Fig. 2 Comparison of the trends for non-modified and modified AMMaxT series at the Tuole station图2 托勒站AMMaxT序列趋势变异修正前后对比 |
Fig. 3 Comparison of the change point variations for the non-modified and modified AMMaxT series at the Yeniugou station图3 野牛沟站AMMaxT序列跳跃变异修正前后对比 |
Fig. 4 Results of K-S test for extreme high temperature indices in the Heihe River Basin图4 黑河流域极端高温指数K-S检验结果 |
Tab. 4 Sum of the ranked goodness of fitting for different distribution models表4 不同理论概率分布模型的拟合优度排名总和 |
| 理论概率分布模型 | AMMaxT | HI |
|---|---|---|
| Beta | 37 | 33 |
| Gamma | 47 | - |
| GEV | 19 | 18 |
| GPD | 51 | 42 |
| Log-Logistic | 62 | - |
| Lognormal | 52 | - |
| Wakeby | 20 | 16 |
| Weibull | 36 | - |
Tab. 5 Estimated values of AMMaxT series at different return periods based on theGEV distribution in the Heihe River Basin表5 基于GEV分布不同重现期水平下黑河流域AMMaxT估算值 |
| 站点 | 10年一遇 | 20年一遇 | 50年一遇 | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 修正前(℃) | 修正后(℃) | 变化率 (%) | 修正前(℃) | 修正后(℃) | 变化率 (%) | 修正前(℃) | 修正后(℃) | 变化率 (%) | ||
| 上游 | 托勒 | 7.55 | 6.95 | 8.06 | 7.78 | 7.03 | 9.62 | 8.02 | 7.11 | 11.33 |
| 野牛沟 | 7.49 | 7.27 | 2.99 | 7.73 | 7.40 | 4.33 | 7.98 | 7.51 | 5.86 | |
| 祁连 | 11.14 | 10.84 | 2.68 | 11.35 | 10.91 | 3.90 | 11.58 | 10.97 | 5.26 | |
| 中游 | 山丹 | 15.66 | 15.45 | 1.40 | 15.92 | 15.65 | 1.67 | 16.19 | 15.88 | 1.88 |
| 张掖 | 16.84 | 16.55 | 1.75 | 17.10 | 16.70 | 2.39 | 17.38 | 16.84 | 3.12 | |
| 酒泉 | 15.83 | 15.59 | 1.52 | 16.05 | 15.71 | 2.11 | 16.26 | 15.81 | 2.76 | |
| 高台 | 16.93 | 16.66 | 1.63 | 17.18 | 16.79 | 2.25 | 17.43 | 16.92 | 2.95 | |
| 下游 | 鼎新 | 17.53 | 17.29 | 1.36 | 17.71 | 17.40 | 1.77 | 17.87 | 17.49 | 2.17 |
| 额济纳旗 | 17.84 | 17.69 | 0.79 | 18.00 | 17.86 | 0.76 | 18.13 | 18.01 | 0.67 | |
注:加粗字体表示同一重现期水平下估算值中较大的数值。 |
Tab. 6 Estimated values of HI series at different return periods based on the GEV distribution in the Heihe River Basin表6 基于GEV分布不同重现期水平下黑河流域HI估算值 |
| 站点 | 10年一遇 | 20年一遇 | 50年一遇 | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 修正前(℃) | 修正后(℃) | 变化率 (%) | 修正前(℃) | 修正后(℃) | 变化率 (%) | 修正前(℃) | 修正后(℃) | 变化率(%) | ||
| 上游 | 托勒 | 1.88 | 1.78 | 5.01 | 2.08 | 1.97 | 5.28 | 2.30 | 2.18 | 5.28 |
| 野牛沟 | 1.72 | 1.62 | 5.67 | 1.95 | 1.78 | 8.63 | 2.24 | 1.96 | 12.22 | |
| 祁连 | 2.04 | 2.00 | 2.40 | 2.27 | 2.20 | 2.80 | 2.52 | 2.44 | 3.17 | |
| 中游 | 山丹 | 2.60 | 2.50 | 3.72 | 2.80 | 2.67 | 4.58 | 3.00 | 2.84 | 5.42 |
| 张掖 | 2.76 | 2.49 | 9.57 | 3.14 | 2.66 | 15.11 | 3.60 | 2.83 | 21.44 | |
| 酒泉 | 2.76 | 2.43 | 11.85 | 3.08 | 2.58 | 16.37 | 3.45 | 2.71 | 21.21 | |
| 高台 | 2.87 | 3.12 | -8.49 | 3.16 | 3.31 | -4.53 | 3.47 | 3.48 | -0.38 | |
| 下游 | 鼎新 | 2.89 | 2.70 | 6.48 | 3.23 | 2.90 | 9.97 | 3.61 | 3.11 | 13.99 |
| 额济纳旗 | 3.30 | 3.04 | 7.96 | 3.81 | 3.38 | 11.35 | 4.45 | 3.77 | 15.23 | |
注:加粗字体表示同一重现期水平下估算值中较大的数值。 |
Fig. 5 Comparison of return periods for non-modified and modified extreme high temperature series at the Tuole station图5 托勒站极端高温序列修正前后重现期对比 |
The authors have declared that no competing interests exist.
| [1] |
|
| [2] |
|
| [3] |
|
| [4] |
[
|
| [5] |
[
|
| [6] |
[
|
| [7] |
[
|
| [8] |
[
|
| [9] |
[
|
| [10] |
[
|
| [11] |
[
|
| [12] |
[
|
| [13] |
[
|
| [14] |
[
|
| [15] |
[
|
| [16] |
[
|
| [17] |
[
|
| [18] |
[
|
| [19] |
[
|
| [20] |
|
| [21] |
|
| [22] |
[
|
| [23] |
|
| [24] |
[
|
| [25] |
[
|
| [26] |
[
|
| [27] |
[
|
| [28] |
[
|
| [29] |
[
|
| [30] |
[
|
| [31] |
[
|
| [32] |
[
|
/
| 〈 |
|
〉 |