欢迎访问《农学学报》，
农学学报

农学学报 ›› 2024, Vol. 14 ›› Issue (10): 53-60.doi: 10.11923/j.issn.2095-4050.cjas2023-0193

所属专题： 水产渔业

• 畜牧 兽医 水产 • 上一篇    下一篇

基于气象因子的河蟹产量估算

吴芳1(), 章雯2, 翟晓瑶3, 龚佳1, 张自强1, 袁昌洪2()   

  1. 1 兴化市气象局，江苏兴化 225700
    2 泰州市气象局，江苏泰州 225300
    3 湖州市南浔区气象局，浙江湖州 313001
  • 收稿日期:2023-09-06 修回日期:2023-11-28 出版日期:2024-10-20 发布日期:2024-10-17
  • 通讯作者:
    袁昌洪，男，1971年出生，江苏姜堰人，正高级工程师，博士，研究方向：应用气象服务技术。通信地址：225300 江苏省泰州市海陵南路299号 泰州市气象局，E-mail：。
  • 作者简介:

    吴芳，女，1992年出生，江苏兴化人，工程师，硕士，研究方向：农业气象。通信地址：225700 江苏省泰州市兴化市垛田街道中和路71号 兴化市气象局，Tel：0523-83262396，E-mail：。

  • 基金资助:
    江苏省气象局指导性项目“兴化市大闸蟹气候品质评价指标及模型研究”(ZD202425); 泰州市气象局科研项目“基于气象条件下兴化市河蟹产量估算模型研究”(202205)

Estimation of Crab Yield Based on Meteorological Factors

WU Fang1(), ZHANG Wen2, ZHAI Xiaoyao3, GONG Jia1, ZHANG Ziqiang1, YUAN Changhong2()   

  1. 1 Xinghua Meteorological Administration, Xinghua 225700, Jiangsu, China
    2 Taizhou Meteorological Administration, Taizhou 225300, Jiangsu, China
    3 Nanxun Meteorological Administration, Huzhou 313001, Zhejiang, China
  • Received:2023-09-06 Revised:2023-11-28 Online:2024-10-20 Published:2024-10-17

RichHTML

0

PDF (PC)

10

摘要：

为探索基于气象因子的河蟹产量估算方法，现利用2010—2022年兴化市河蟹产量和同期气象数据，与3种机器学习算法相结合，在5个关键生长阶段（放养期、第1~3次脱壳期、第4次脱壳期、第5次脱壳期和成熟捕捞期），构建河蟹产量估算模型。结果表明，在河蟹不同生长阶段，不同气象因子估算河蟹产量的精度差异显著。总体来说，基于支持向量机回归(SVR)算法所构建的河蟹产量估算模型要优于随机森林回归(RFR)算法和偏最小二乘回归(PLSR)算法，预测精度R2在0.95~0.98，RMSE在24.07~35.14 kg/hm2之间变化。在单一生长阶段中，在放养期的估算精度最高，R2为0.98，RMSE为24.07 kg/hm2，在全生育期中模型预测精度R2为0.92，RMSE为44.31 kg/hm2。研究结果揭示了气象因子在不同生长阶段下估算河蟹产量的能力。

关键词: 河蟹产量, 气象因子, 随机森林回归, 支持向量回归, 偏最小二乘回归

Abstract:

This study aims to estimate crab yield by using meteorological factors. The crab yield and meteorological factors were acquired at five key growth stages (stocking period, 1st-3rd shelling period, 4th shelling period, 5th shelling period and mature fishing period) from 2010-2022 in Xinghua of Jiangsu. Meanwhile, the meteorological factors were used as the inputs for the three machine learning algorithms for the crab yield estimation. The three machine learning algorithms included the Random Forest Regression (RFR), Support Vector Regression (SVR), and Partial Least Squares Regression (PLSR). Finally, the estimation models were developed using Leave-One-Out cross validation under different growth stages. The results showed that the precision of estimating crab yield by different meteorological factors was significantly different at different growth stages. For the three machine learning algorithms, the SVR was found to be superior to the RFR and the PLSR (coefficient of determination, R2=0.95-0.98, RMSE=24.07-35.14 kg/hm2). In the single growth stage, the highest estimation accuracy was observed at the stocking period (R2=0.98, RMSE=24.07 kg/hm2). In the whole growth period, the estimation accuracy was (R2=0.92, RMSE=44.31 kg/hm2). Therefore, the meteorological factors can be expected to estimate crab yield under the different growth stages.

Key words: crab yield, meteorological factors, RFR, SVR, PLSR