Modelling soil erodibility in mountain rangelands of southern Kyrgyzstan

Soil erosion in mountain rangelands in Kyrgyzstan is an emerging problem due to vegetation loss caused by overgrazing. It is further exacerbated by mountain terrain and high precipitation values in Fergana range in the south of Kyrgyzstan. The main objective of this study was to map soil erodibility in the mountainous rangelands of Kyrgyzstan. The results of this effort are expected to contribute to the development of soil erodibility modelling approaches for mountainous areas. In this study, we mapped soil erodibility at two sites, both representing grazing rangelands in the mountains of Kyrgyzstan and having potentially different levels of grazing pressure. We collected a total of 232 soil samples evenly distributed in geographical space and feature space. Then we analyzed the samples in laboratory for grain size distribution and calculated soil erodibility values from these data using the Revised Universal Soil Loss Equation (RUSLE) K-factor formula. After that, we derived different terrain indices and ratios of frequency bands from ASTER GDEM and LANDSAT images to use as auxiliary data because they are among the main soil forming factors and widely used for prediction of various soil properties. Soil erodibility was significantly correlated with channel network base level (geographically extrapolated altitude of water channels), remotely sensed indices of short-wave infrared spectral bands, exposition, and slope degree. We applied multiple regression analysis to predict soil erodibility from spatially explicit terrain and remotely sensed indices. The final soil erodibility model was developed using the spatially explicit predictors and the regression equation and then improved by adding the residuals. The spatial resolution of the model was 30 m, and the estimated mean adjusted coefficient of determination was 0.47. The two sites indicated different estimated and predicted means of soil erodibility values (0.035 and 0.039) with a 0.05 significance level, which is attributed mainly to the considerable difference in elevation.     

采用卷积神经网络构建西北太平洋柔鱼渔场预报模型

对远洋渔场资源和位置进行预报可以为远洋渔业生产及管理提供重要信息。该研究针对西北太平洋柔鱼渔场，利用海洋表面温度遥感信息和中国远洋渔船生产资料，基于深度学习原理，选取卷积神经网络构建西北太平洋柔鱼渔场预报模型。根据不同月份、不同通道构建了多种数据集，用于训练渔场预报模型。训练结果表明，4个通道组合的数据集的训练结果最优，渔汛早期（7-8月）、中期（9月）和后期（10-11月）测试样本的准确率分别为80.5%、81.5%和81.4%。以2015年的真实渔场数据对模型进行验证，模型的平均召回率为82.3%，平均精确率为66.6%，F1得分平均值为73.1%，预测的高产渔区与实际作业的高单位捕捞努力量渔获量区基本匹配。该研究构建的渔场预报模型可以获得较好的准确率，可为其他鱼种的渔场预报模型构建提供新的思路。     

Deep neural network algorithm for estimating maize biomass based on simulated Sentinel 2A vegetation indices and leaf area index

Accurate estimation of biomass is necessary for evaluating crop growth and predicting crop yield.Biomass is also a key trait in increasing grain yield by crop breeding.The aims of this study were(i)to identify the best vegetation indices for estimating maize biomass,(ii)to investigate the relationship between biomass and leaf area index(LAI)at several growth stages,and(iii)to evaluate a biomass model using measured vegetation indices or simulated vegetation indices of Sentinel 2A and LAI using a deep neural network(DNN)algorithm.The results showed that biomass was associated with all vegetation indices.The three-band water index(TBWI)was the best vegetation index for estimating biomass and the corresponding R2,RMSE,and RRMSE were 0.76,2.84 t ha−1,and 38.22%respectively.LAI was highly correlated with biomass(R2=0.89,RMSE=2.27 t ha−1,and RRMSE=30.55%).Estimated biomass based on 15 hyperspectral vegetation indices was in a high agreement with measured biomass using the DNN algorithm(R2=0.83,RMSE=1.96 t ha−1,and RRMSE=26.43%).Biomass estimation accuracy was further increased when LAI was combined with the 15 vegetation indices(R2=0.91,RMSE=1.49 t ha−1,and RRMSE=20.05%).Relationships between the hyperspectral vegetation indices and biomass differed from relationships between simulated Sentinel 2A vegetation indices and biomass.Biomass estimation from the hyperspectral vegetation indices was more accurate than that from the simulated Sentinel 2A vegetation indices(R2=0.87,RMSE=1.84 t ha−1,and RRMSE=24.76%).The DNN algorithm was effective in improving the estimation accuracy of biomass.It provides a guideline for estimating biomass of maize using remote sensing technology and the DNN algorithm in this region.     

基于改进极限学习机的水体溶解氧预测方法

为了有效地指导水产养殖生产，提高溶解氧浓度预测的精度，提出了基于因子筛选和改进极限学习机（Extreme Learning Machine，ELM）的水产养殖溶解氧预测模型。首先，利用皮尔森相关系数法计算各影响因子与溶解氧浓度间的相关系数，提取强关联因子，降低预测模型的输入量维度；采用偏最小二乘算法（Partial Least Square, PLS）优化传统ELM神经网络，避免网络中隐含层共线性问题，保障输出权值的稳定性；然后，结合新型激活函数，构建水体溶解氧浓度预测模型。最后，将SPLS-ELM（Selection Based Partial Least Square Optimized Extreme Learning Machine）预测模型应用到江苏省无锡市南泉基地某试验池塘的水体溶解氧预测中。试验结果表明：该模型的预测均方根误差为0.3232，与最小二乘支持向量机（Least Square Support Vector Machine，LSSVM）、BP神经网络、粒子群（Particle Swarm Optimization，PSO）优化LSSVM和遗传算法（Genetic Algorithm, GA）优化BP神经网络相比分别降低40.98%、44.48%、34.73%和44.18%。且该模型的运行时间仅0.6231s，预测精度和运行效率明显优于其他模型。该模型的溶解氧预测曲线接近真实溶解氧变化曲线，能够满足水产养殖实际生产对水体溶解氧预测的要求。     

基于深度学习的大豆生长期叶片缺素症状检测方法

为了检测作物叶片缺素,提出了一种基于神经网络的大豆叶片缺素视觉检测方法。在对大豆缺素叶片进行特征分析后,采用深度学习技术,利用Mask R-CNN模型对固定摄像头采集的叶片图像进行分割,以去除背景特征,并利用VGG16模型进行缺素分类。首先通过摄像头采集水培大豆叶片图像,对大豆叶片图像进行人工标记,建立大豆叶片图像分割任务的训练集和测试集,通过预训练确定模型的初始参数,并使用较低的学习率训练Mask RCNN模型,训练后的模型在测试集上对背景遮挡的大豆单叶片和多叶片分割的马修斯相关系数分别达到了0.847和0.788。通过预训练确定模型的初始参数,使用训练全连接层的方法训练VGG16模型,训练的模型在测试集上的分类准确率为89.42%。通过将特征明显的叶片归类为两类缺氮特征和4类缺磷特征,分析讨论了模型的不足之处。本文算法检测一幅100万像素的图像平均运行时间为0.8 s,且对复杂背景下大豆叶片缺素分类有较好的检测效果,可为农业自动化生产中植株缺素情况估计提供技术支持。     

网络对农业高校大学生学习的影响及对策

网络现已成为大学生日常生活中不可或缺的一部分。网络所具有的资源丰富、快速便捷等特点给大学生带来了很多积极影响,但是网络所带来的消极影响也不容小觑。通过对农业院校大学生学习情况的调查,研究并总结了网络对大学生学习的积极和消极影响。针对网络对大学生学习的负面影响提出了几点应对之策。     

基于产品空间理论的中国林业产业升级特征分析

林业产业如何进行转型升级是实现林业现代化与高质量发展中的重要一环。本文基于产品空间理论，使用聚类热图和复杂网络方法，构建1962—2017年中国林业产业产品空间，从产品复杂度和优势林产品变化两个角度分析林业产业升级特征。结果表明，新获得优势林产品复杂度高，为0.31，传统失势林产品较低，为-0.21，表征林业产业已朝附加值更高的有利方向升级，但传统优势林产品复杂度较低，说明林业产业已有优势产品竞争激烈不利于升级。产品复杂度较高的潜在优势林产品是林业机械、机械零件和林化产品，优势林产品在产品空间中的变化存在一定路径依赖，先扩散后收敛于产品空间的中心位置，政府制定的林业产业政策与全球化背景可能促进了优势林产品升级。因此，政府制定产业政策时，应对不同优势类型林产品采取差异化策略，应完善产品复杂度低的传统优势林产品市场退出机制；减少与优势林产品距离较近的潜在优势林产品政策干预，对距离较远的复杂度高的潜在优势林产品使用补贴等方式引导升级。     

适应农业产业发展产教融合课堂教学改革研究

本文对适应农业区域产业发展产教融合课堂教学改革进行了一定的研究，希望可以为“产教 结合、校企一体”的发展之路提供一定的参考价值。     

采用全卷积神经网络与Stacking算法的湿地分类方法

高精度湿地制图对湿地生态保护与精细管理具有重要的支撑作用。针对传统湿地分类方法的精度不高和泛化能力弱等问题，提出了一种联合全卷积神经网路（fully convolutional neural network，FCN）与集成学习的湿地分类方法。首先利用全卷积神经网络（SegNet、UNet及RefineNet）对GF-6影像的语义特征进行提取与融合，然后利用Stacking集成算法对融合后的特征进行判别和分类。结果表明，联合全卷积神经网络与Stacking算法能有效提取湿地信息，总体分类精度为88.16%，Kappa系数为0.85。与联合全卷积神经网络与单一机器学习RF、SVM与kNN算法相比，该文提出的湿地分类方法在总体分类精度上分别提高了4.87%，5.31%和5.08%；与联合单一全卷积神经网络（RefineNet、SegNet、UNet）与Stacking算法下的湿地分类结果，该文提出的湿地分类方法在总体分类精度上分别提高了2.78%，4.48%与4.91%；该文方法一方面能通过卷积神经网络提取遥感影像深层的语义特征，另一方面通过集成学习根据各分类器的表征性能进行合理的选择并重组，从而提高分类精度及其泛化能力。该方法能为湿地信息提取及土地覆盖分类方法的研究提供参考。     

花龄期棉花虫害的电子鼻检测

棉花害虫具有隐蔽性、迁飞性和突发性特点，并且影响因素众多，棉花虫害准确地诊断是农业领域的难点问题。该研究以受到棉铃虫侵害的花铃期棉花为研究对象，采用电子鼻对不同处理的棉花挥发物进行检测。研究表明，主成分分析（Principal Component Analysis, PCA）和聚类分析结果显示健康棉花释放的挥发物具有明显的昼夜节律性，健康棉花与虫害棉花差异性显著。径向基函数神经网络（Radial Basis Function Neural Network, RBFNN）对8个不同时间的4组虫害棉花处理进行分析，测试集判别总的正确率为73.4%，健康棉花对照组测试集判别正确率100%，误判样本出现在3个虫害处理之间。当不考虑时间因素建立虫害棉花统一的预测模型，RBFNN模型对健康棉花对照组的预测正确率均达到了100%，分析结果可以作为花铃期棉花是否遭受棉铃虫侵害的依据，说明电子鼻可以作为棉花虫害发生的有效监测手段，在农作物虫害监测领域具有潜在的应用价值。