Time series modelling of increased soil temperature anomalies during long period
More details
Hide details
Department of Water Engineering, Oceanic and Atmospheric Research Centre, College of Agriculture, Shiraz University, Shiraz, Iran
Department of Soil Science, College of Agriculture, Shiraz University, Shiraz, Iran
Publication date: 2015-10-28
Int. Agrophys. 2015, 29(4): 509–515
Soil temperature just beneath the soil surface is highly dynamic and has a direct impact on plant seed germination and is probably the most distinct and recognisable factor governing emergence. Autoregressive integrated moving average as a stochastic model was developed to predict the weekly soil temperature anomalies at 10 cm depth, one of the most important soil parameters. The weekly soil temperature anomalies for the periods of January1986-December 2011 and January 2012-December 2013 were taken into consideration to construct and test autoregressive integrated moving average models. The proposed model autoregressive integrated moving average (2,1,1) had a minimum value of Akaike information criterion and its estimated coefficients were different from zero at 5% significance level. The prediction of the weekly soil temperature anomalies during the test period using this proposed model indicated a high correlation coefficient between the observed and predicted data – that was 0.99 for lead time 1 week. Linear trend analysis indicated that the soil temperature anomalies warmed up significantly by 1.8°C during the period of 1986-2011.