TY - JOUR
T1 - Uncertainty quantification of nitrogen use efficiency prediction in China using Monte Carlo simulation and quantile regression forests
AU - Liu, Yingxia
AU - Heuvelink, Gerard B.M.
AU - Bai, Zhanguo
AU - He, Ping
PY - 2023/1
Y1 - 2023/1
N2 - Nitrogen use efficiency (NUE) plays an essential role in food security and environmental sustainability. With the development of technology, NUE prediction by models has come available. However, the prediction uncertainty of NUE models is still poorly understood. This study aimed to analyze uncertainty in NUE predictions obtained from a random forest machine learning model. Input and model uncertainties were quantified using Monte Carlo simulation in three scenarios and quantile regression forests (QRF), respectively, to analyze how these uncertainties propagate to the NUE predictions for 31 provinces in China from 1978 to 2015. Two NUE indicators were considered: the partial factor productivity of nitrogen (PFPN) and the partial nutrient balance of nitrogen (PNBN). The results indicated that the prediction uncertainty for both NUE indicators decreased over time. In 2015, PFPN had a higher 90% prediction interval ratio (PIR90) of input data in south and west China and a higher 90% prediction interval width (PIW90) in south and east-coastal China, while PNBN had a higher PIR90 in north China and a higher PIW90 in northeast China. The NUE prediction uncertainty propagated from QRF models had similar spatial patterns as those resulting from uncertainty in input data. NUE in most provinces had smaller input uncertainty than model uncertainty, except PNBN, which had smaller model uncertainty than input uncertainty after 2010. Generally, PNBN had higher input uncertainty contributions than PFPN in 2015, especially in south and northeast China. Overall, the uncertainties in NUE predictions were substantial. A series of recommendations were made to improve the accuracy of NUE predictions. These may be applied by the government, in order to inform sustainable nitrogen management in agroecological systems.
AB - Nitrogen use efficiency (NUE) plays an essential role in food security and environmental sustainability. With the development of technology, NUE prediction by models has come available. However, the prediction uncertainty of NUE models is still poorly understood. This study aimed to analyze uncertainty in NUE predictions obtained from a random forest machine learning model. Input and model uncertainties were quantified using Monte Carlo simulation in three scenarios and quantile regression forests (QRF), respectively, to analyze how these uncertainties propagate to the NUE predictions for 31 provinces in China from 1978 to 2015. Two NUE indicators were considered: the partial factor productivity of nitrogen (PFPN) and the partial nutrient balance of nitrogen (PNBN). The results indicated that the prediction uncertainty for both NUE indicators decreased over time. In 2015, PFPN had a higher 90% prediction interval ratio (PIR90) of input data in south and west China and a higher 90% prediction interval width (PIW90) in south and east-coastal China, while PNBN had a higher PIR90 in north China and a higher PIW90 in northeast China. The NUE prediction uncertainty propagated from QRF models had similar spatial patterns as those resulting from uncertainty in input data. NUE in most provinces had smaller input uncertainty than model uncertainty, except PNBN, which had smaller model uncertainty than input uncertainty after 2010. Generally, PNBN had higher input uncertainty contributions than PFPN in 2015, especially in south and northeast China. Overall, the uncertainties in NUE predictions were substantial. A series of recommendations were made to improve the accuracy of NUE predictions. These may be applied by the government, in order to inform sustainable nitrogen management in agroecological systems.
KW - Input data
KW - Model uncertainty
KW - Monte Carlo
KW - Nitrogen use efficiency
KW - Quantile regression forests
U2 - 10.1016/j.compag.2022.107533
DO - 10.1016/j.compag.2022.107533
M3 - Article
AN - SCOPUS:85143684269
SN - 0168-1699
VL - 204
JO - Computers and Electronics in Agriculture
JF - Computers and Electronics in Agriculture
M1 - 107533
ER -