• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Theoretical and applied climatology >Drought forecasting in Luanhe River basin involving climatic indices
【24h】

Drought forecasting in Luanhe River basin involving climatic indices

机译：涉及气候指数的Lu河流域干旱预报

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Drought is regarded as one of the most severe natural disasters globally. This is especially the case in Tianjin City, Northern China, where drought can affect economic development and people's livelihoods. Drought forecasting, the basis of drought management, is an important mitigation strategy. In this paper, we evolve a probabilistic forecasting model, which forecasts transition probabilities from a current Standardized Precipitation Index (SPI) value to a future SPI class, based on conditional distribution of multivariate normal distribution to involve two large-scale climatic indices at the same time, and apply the forecasting model to 26 rain gauges in the Luanhe River basin in North China. The establishment of the model and the derivation of the SPI are based on the hypothesis of aggregated monthly precipitation that is normally distributed. Pearson correlation and Shapiro-Wilk normality tests are used to select appropriate SPI time scale and large-scale climatic indices. Findings indicated that longer-term aggregated monthly precipitation, in general, was more likely to be considered normally distributed and forecasting models should be applied to each gauge, respectively, rather than to the whole basin. Taking Liying Gauge as an example, we illustrate the impact of the SPI time scale and lead time on transition probabilities. Then, the controlled climatic indices of every gauge are selected by Pearson correlation test and the multivariate normality of SPI, corresponding climatic indices for current month and SPI 1, 2, and 3 months later are demonstrated using Shapiro-Wilk normality test. Subsequently, we illustrate the impact of large-scale oceanic-atmospheric circulation patterns on transition probabilities. Finally, we use a score method to evaluate and compare the performance of the three forecasting models and compare them with two traditional models which forecast transition probabilities from a current to a future SPI class. The results show that the three proposed models outperform the two traditional models and involving large-scale climatic indices can improve the forecasting accuracy.
机译：干旱被认为是全球最严重的自然灾害之一。在中国北方的天津市尤其如此，那里的干旱会影响经济发展和民生。干旱预报是干旱管理的基础，是一项重要的缓解策略。在本文中，我们演化了一个概率预测模型，该模型基于多变量正态分布的条件分布，同时涉及两个大规模气候指数，预测了从当前标准降水指数（SPI）值到未来SPI类的过渡概率。时间，并将预测模型应用于华北the河流域的26个雨量器。该模型的建立和SPI的推导基于正态分布的总月降水量假设。使用Pearson相关和Shapiro-Wilk正态性检验来选择适当的SPI时间尺度和大规模气候指数。研究结果表明，一般而言，长期累积的月降水量更可能被认为是正态分布的，因此应将预报模型分别应用于每个雨量器而不是整个流域。以Liying Gauge为例，我们说明了SPI时间尺度和提前期对过渡概率的影响。然后，通过皮尔逊相关检验选择每个雨量器的受控气候指数，并用SPI的多元正态性，使用Shapiro-Wilk正态性检验证明当月和SPI 1、2和3个月的相应气候指数。随后，我们说明了大规模海洋-大气环流模式对过渡概率的影响。最后，我们使用评分方法评估和比较这三种预测模型的性能，并将它们与两种传统模型进行比较，这两种模型预测从当前SPI类到未来SPI类的转换概率。结果表明，所提出的三个模型优于两个传统模型，并且涉及大规模的气候指标可以提高预报的准确性。

著录项

  • 来源
    《Theoretical and applied climatology》 |2017年第4期|1133-1148|共16页
  • 作者

    Ren Weinan; Wang Yixuan; Li Jianzhu; Feng Ping; Smith Ronald J.;

  • 作者单位

    Tianjin Univ, State Key Lab Hydraul Engn Simulat & Safety, Tianjin 300072, Peoples R China;

    Tianjin Univ, State Key Lab Hydraul Engn Simulat & Safety, Tianjin 300072, Peoples R China;

    Tianjin Univ, State Key Lab Hydraul Engn Simulat & Safety, Tianjin 300072, Peoples R China;

    Tianjin Univ, State Key Lab Hydraul Engn Simulat & Safety, Tianjin 300072, Peoples R China;

    Purdue Univ, Agr & Biol Engn Dept, 225 South Univ St, W Lafayette, IN 47907 USA|Purdue Univ, ESE IGP, 225 South Univ St, W Lafayette, IN 47907 USA;

  • 收录信息
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

    Drought forecasting; Transition probabilities; SPI; Large-scale climatic indices;

    机译：干旱预报过渡概率SPI大气候指数;

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号