基于遥感信息和净初级生产力的天然橡胶估产模型

Estimation Model of Natural Rubber Yield Based on Net Primary Production and Remote Sensing

建立基于MODIS数据的天然橡胶净初级生产力遥感估算模型,利用2009年生长季(4—12月)250 m分辨率的MODIS数据和气象数据估算海南阳江农场天然橡胶的生长季净初级生产力,通过天然橡胶的干物质分配率估算阳江农场天然橡胶的产胶潜力。用以树位为单元的地面实际干胶产量与已得到的天然橡胶林产胶潜力进行回归分析,建立天然橡胶干胶产量的估测模型。利用估产模型对阳江农场2010年7月的干胶产量进行模拟,用同期实际干胶产量对估产模型进行精度验证与实用性评价,结果显示7月估产模型均方根误差RMSE为1.78 g.m-2,相对均方根误差RMSEr为18.25%。研究表明,基于遥感信息和净初级生产力的天然橡胶估产模型具有较好的产量估测效果。

A net primary productivity（NPP） model for natural rubber was established with 250 m×250 m MODIS remote images and meteorological data on the Yangjiang Farm at growing season from April to December 2009.And the potential productivity of natural rubber was estimated based on the NPP and the natural rubber allocation rate of dry matter.The natural rubber yield estimation model was established with actual rubber yield in each task and the potential productivity of natural rubber based on regression analysis.The accuracy of yield estimation model was validated by the actual rubber production data in July 2010 on the Yangjiang Farm.The root mean square error（RMSE） is 1.78 g·m-2,and the relative root mean square error（RMSEr）is 18.25%.The results show that the natural rubber yield estimation model based on remote sensing information and net primary productivity has a good effect.