|
|
|
| Accuracy assessment of rainfall forecast and analysis on water-saving irrigation strategies during rice growing seasons in South China |
| LUO Yufeng1,2, MA Zhen1, LYU Xinwei3, GUO Longzhu1, CUI Yuanlai2, XIANG Zhao2 |
| 1.State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing, Jiangsu 210098, China; 2.State Key Laboratory of Water Resources and Hydropower Engineering Science, Wuhan University, Wuhan, Hubei 430072, China; 3. School of Software Engineering, University of Science and Technology of China, Hefei, Anhui 230051, China |
|
|
|
|
Abstract Increasing effective rainfall is one of strategies for saving irrigation water, unfortunately there is an uncertainty in future rainfall, thus accuracy assessment of rainfall forecast is necessary for effective rainfall use. Taking the Guanfu Plain Irrigation District as an example, we collected 7-day lead time rainfall forecast data in three seasons of early and late rice(2012—2015)from the National Meteorological Centre of China, and the observed rainfall data, then analyzed the forecast accuracy by using correct forecast rate, threat score(TS), false alarm rate(FAR), missing alarm rate(MAR)and receiver operating characteristic(ROC)curve. Furthermore, strategies for increasing effective rainfall are proposed based on the accuracy assessment. The results show that the correct forecast percentage of rain/no rain during the rice growing seasons has reached 76.3%, the MAR is lower but the FAR is higher in the early rice season than in the late rice season. The difference in TS for light rain during the early and late rice seasons is not significant; for middle and above rainfalls, however, the TS in the late season is lower than in the early season. The areas under the ROC are greater than 0.5 for all 1-7 days lead time, implying that the correct forecast percentage is higher than the FAR and all the forecasts are positive, thus the correct forecast percentage can be used in irrigation management. The correct forecast percentage is ranged from 80.1% to 88.6% in the re-greening, milk-maturing and yellow-maturing stages of early rice. Thus if a light rain or above is forecasted, then there should be no or less irrigation. The correct forecast percentage is slightly lower and ranged from 64.9% to 77.2% in the early tillering, late tillering, jointing/booting, and heading/flowering stages. Therefore, if a continuous clear day or slight rain is forecasted, then irrigation should be applied according to the crop requirement. However, if a medium or above rainfall in 2-3 days is forecasted, then irrigation amount should be reduced. Although the accuracy in the early rice season is lower than in the late rice season, the forecasted and actual rainfall events are even more in the early rice season, so that the irrigation amount can be decreased if a medium or above rainfall is forecasted. After a decision on irrigation amount is made by taking an account of rain forecast accuracy, not only natural rainfall will be utilized sufficiently but also water wasting can be avoided to achieve the water saving goal.
|
|
Received: 11 December 2015
|
|
|
|
| [1]茆智. 水稻节水灌溉[J]. 中国农村水利水电, 1997(4):45-47. MAO Zhi. Water saving irrigation of rice[J]. China rural water and hydropower, 1997(4):45-47.(in Chinese)[2]徐凤琴. 有效降水量浅析[J]. 气象水文海洋仪器, 2009(1): 96-100. XU Fengqin. Concise analysis of effextive precipitation[J]. Meteorological, hydrological and marine instruments, 2009(1):96-100.(in Chinese)[3]吕露, 冯常萍, 崔远来. 水稻不同节水灌溉模式的水分利用效率比较——以荆门、桂林为例[J]. 节水灌溉,2011(3): 15-17, 22. LYU Lu, FENG Changping, CUI Yuanlai. Comparison of water-using efficiency for paddy rice under different water-saving irrigation methods[J]. Water saving irrigation, 2011(3): 15-17, 22.(in Chinese)[4]林明智, 毕宝贵, 乔林. 中央气象台短期降雨预报水平初步分析[J]. 应用气象学报, 1995,6(4): 392-399. LIN Mingzhi, BI Baogui, QIAO Lin. A preliminary analysis of short-range forecasting level for precipitation by Central Meteorological Office of China[J]. Quarterly journal of applied meteorology, 1995,6(4): 392-399.(in Chinese)[5]ROCHESTER E W, BUSCH C D. An irrigation scheduling model which incorporates rainfall predictions[J]. Journal of the American Water Resources Association, 1972,8(3): 608-613.[6]WANG Dingbao, CAI Ximing. Irrigation scheduling—role of weather forecasting and farmers′ behavior[J]. Journal of water resources planning and management, 2009, 135(5): 364-372.[7]AGNIHOTRI G, MOHAPATRA M. Prediction of occurrence of daily summer monsoon precipitation over Karnataka[J]. Meteorological applications, 2012,19(2): 130-139.[8]MISHRA A, SIDERIUS C, ABERSON K, et al. Short-term rainfall forecasts as a soft adaptation to climate change in irrigation management in North-East India[J]. Agricultural water management, 2013,127: 97-106.[9]GOWING J W, EJIEJI C J. Real-time scheduling of supplemental irrigation for potatoes using a decision model and short-term weather forecasts[J]. Agricultural water management, 2001,47(2):137-153.[10]侯静文,罗玉峰,崔远来. 降雨预报准确度分析及其在水稻节水灌溉决策中的应用[J]. 节水灌溉,2013(3):24-26,30. HOU Jingwen, LUO Yufeng, CUI Yuanlai. Analysis of rainfall forecast accuracy and its application in water-saving irrigation decision-making for rice[J]. Water saving irrigation, 2013(3):24-26,30.(in Chinese)[11]张星星, 朱成立, 彭世彰, 等. 降水预报准确度分析及其在提高降水利用率中的作用[J]. 河海大学学报(自然科学版), 2014, 42(3): 230-233. ZHANG Xingxing, ZHU Chengli, PENG Shizhang, et al. Accuracy analysis of rainfall forecast and its application to improvement of utilization rate of rainfall[J]. Journal of Hohai University(natural sciences), 2014, 42(3): 230-233.(in Chinese)[12]詹道江,叶守泽. 工程水文学[M]. 3版. 北京:中国水利水电出版社,2000.[13]中国气象局.地面气象观测规范: QX/T 52—2007[S]. 北京:气象出版社,2007.[14]WILKS D S. Statistical methods in the atmospheric sciences[M]. 2nd ed.San Diego:Academic press, 2005.[15]ZHANG Xiaoli, LI Xiongfei, FENG Yuncong, et al. The use of ROC and AUC in the validation of objective image fusion evaluation metrics[J]. Signal processing, 2015, 115: 38-48. |
|
|
|
