Prediction of the livestock carrying capacity using neural network in the meadow steppe

Abstract

In order to predict the livestock carrying capacity in meadow steppe, a method using back propagation neural network is proposed based on the meteorological data and the remote-sensing data of Normalised Difference Vegetation Index. In the proposed method, back propagation neural network was first utilised to build a behavioural model to forecast precipitation during the grass-growing season (June–July–August) from 1961 to 2015. Second, the relationship between precipitation and Normalised Difference Vegetation Index during the grass-growing season from 2000 to 2015 was modelled with the help of back propagation neural network. The prediction results demonstrate that the proposed back propagation neural network-based model is effective in the forecast of precipitation and Normalised Difference Vegetation Index. Thus, an accurate prediction of livestock carrying capacity is achieved based on the proposed back propagation neural network-based model. In short, this work can be used to improve the utilisation of grassland and prevent the occurrence of vegetation degradation by overgrazing in drought years for arid and semiarid grasslands.

A new temporal prediction method of grazing pressure based on normalized difference vegetation index and precipitation using nonlinear autoregressive with exogenous input networks

“…Consequently, the grazing pressure index I p from 2016 to 2020 can be predicted using the proposed method as shown in Table 3. Several studies have been presented to predict the NDVI with respect to the precipitation such as multiple linear regression (Iwasaki, 2009), SVM (Huang et al, 2017) and BPNN (Wu et al, 2019). This paper has proposed to introduce the NARX network to predict the temporal variations of the NDVI with respect to the precipitation.…”

“…The data of the monthly precipitation during the years from 1975 to 2015 were provided by the China Meteorological Data Sharing Service System (http://data.cma.cn). The mean monthly precipitation and the mean monthly NDVI in the grass‐growing season are represented by the mean value from June to August (Shi et al, ; Wu et al, ). Moreover, the actual numbers of the livestock in the mid‐year from 2000 to 2015 were obtained by the statistical data of Hulunbuir Bureau Agriculture and Animal Husbandry.…”

“…The data of the monthly precipitation during the years from 1975 to 2015 were provided by the China Meteorological Data Sharing Service System (http://data.cma.cn). The mean monthly precipitation and the mean monthly NDVI in the grass-growing season are represented by the mean value from June to August (Shi et al, 2001;Wu et al, 2019).…”

“…The vegetation productivity can be approximately represented by the normalized difference vegetation index (NDVI) with the development of the remote sensing technology (Kawamura et al, 2005;Liang et al, 2012;Reiche et al, 2012). Because the temporal variations of the NDVI is highly correlated with the precipitation in arid and semi-arid grasslands (Richard et al, 2008), the NDVI variations can be predicted through the precipitation (Dulamsuren & Hauck, 2008;Huang, Ming, Huang, Leng, & Hou, 2017;Lin & Zhang, 2013;Wu, Fu, Jin, Wu, & Bai, 2019).…”

“…Studies have been presented to predict the temporal variations of the NDVI with respect to the precipitation, such as the statistical regression methods (Huang et al, 2017), back-propagation neural network (BPNN) (Wu et al, 2019;Yang, Zhu, Zhao, Liu, & Tong, 2011), support vector machine (SVM) (Huang et al, 2017) and so on. The BPNN can accurately predict the NDVI through a certain amount of the time series of the precipitation.…”