Meteorological drought forecasting for ungauged areas based on machine learning: Using long-range climate forecast and remote sensing data
Meteorological drought forecasting for ungauged areas based on machine learning: Using long-range climate forecast and remote sensing data
复制标题
DOI：
10.1016/j.agrformet.2017.02.011
复制
发表时间：
2017-05
影响因子：
6.2
通讯作者：
J. Rhee;J. Im
中科院分区：
文献类型：
--
作者：
J. Rhee;J. Im
本文建立了一个适用于无资料地区的高分辨率干旱预报模型。标准化降水指数（SPI）和标准化降水蒸散指数（SPEI）分别在3、6、9和12个月时间尺度上进行了预报，预报提前期为1-6个月，分辨率为0.05 × 0.05°。对使用长期气候预报数据和使用没有观测数据的时期的气候数据进行了比较。利用基于遥感数据的干旱相关变量的机器学习模型与克里格空间插值进行了比较。两个性能指标，一个是生产者的干旱准确性，定义为在极端，严重和中度干旱类的样本总数在这些类中的正确分类的样本的数量，另一个是用户的干旱准确性，定义为在干旱类中的正确分类的样本的数量在这些类的样本总数分类。其中一种机器学习模型，极端随机树，在大多数情况下表现最好，生产者的准确度高达64%，而空间插值在用户的准确度方面表现更好，高达44%。在本研究中使用的条件下，长期气候预测数据的贡献并不显着，但如果预测技巧得到改善或使用更复杂的降尺度方法，预计将进一步改善。对降水量和平均气温预报误差的模拟减小进行了检验：降水量预报误差的模拟减小对干旱预报有一定的改善作用，而平均气温预报误差的模拟减小对干旱预报的改善作用不大。虽然仍有一些改进的余地，开发的模型可以用于干旱相关的决策在无测量区域。
A high-resolution drought forecast model for ungauged areas was developed in this study. The Standardized Precipitation Index (SPI) and Standardized Precipitation Evapotranspiration Index (SPEI) with 3-, 6-, 9-, and 12-month time scales were forecasted with 1–6-month lead times at 0.05 × 0.05° resolution. The use of long-range climate forecast data was compared to the use of climatological data for periods with no observation data. Machine learning models utilizing drought-related variables based on remote sensing data were compared to the spatial interpolation of Kriging. Two performance measures were used; one is producer’s drought accuracy, defined as the number of correctly classified samples in extreme, severe, and moderate drought classes over the total number of samples in those classes, and the other is user’s drought accuracy, defined as the number of correctly classified samples in drought classes over the total number of samples classified to those classes. One of the machine learning models, extremely randomized trees, performed the best in most cases in terms of producer’s accuracy reaching up to 64%, while spatial interpolation performed better in terms of user’s accuracy up to 44%. The contribution of long-range climate forecast data was not significant under the conditions used in this study, but further improvement is expected if forecast skill is improved or a more sophisticated downscaling method is used. Simulated decreases of forecast error in precipitation and mean temperature were tested: the simulated decrease of forecast error in precipitation improves drought forecast while the decrease of forecast error in mean temperature does not contribute much. Although there is still some room for improvement, the developed model can be used for drought-related decision making in ungauged areas.