Skip to main content

Table 2 Comparison of candidate SARIMA models for number of CCHF cases in Zahedan district, Iran

From: The effect of climate variables on the incidence of Crimean Congo Hemorrhagic Fever (CCHF) in Zahedan, Iran

Model Variables   Lag Estimate SE t P-value
[A] Monthly patients Constant   1.96 0.40 4.81 < 0.001
   MA Lag 1 0.26 0.09 2.75 0.007
    Lag 2 0.26 0.09 2.77 0.006
[B]   MA Lag 1 −0.89 0.04 −18.57 < 0.001
   MA, seasonal Lag 1 −0.34 0.17 −1.95 0.05
[C]   MA Lag 1 −0.87 0.05 − 14.82 < 0.001
   Seasonal difference   1    
   MA, seasonal Lag 1 −0.37 0.19 −1.95 0.054
  Maximum temperature    0.023 0.14 0.16 0.86
  Minimum temperature    0.26 0.24 1.09 0.27
  Mean temperature    −0.24 0.18 −1.33 0.18
  Rain fall    0.02 0.04 0.57 0.56
  Maximum Humidity    −0.01 0.02 − 0.76 0.44
  Minimum Humidity    −0.19 0.13 −1.46 0.14
  Mean Humidity    0.06 0.06 0.92 0.35
  Sunshine    0.02 0.01 −1.33 0.18
  1. [A]: MA (2), [B]: univariate SARIMA (0,1,1) (0,1,1), [C]: multivariate SARIMA (0,1,1) (0,1,1), SARIMA: seasonal auto-regressive integrated moving average
  2. A) AIC: 4.84, AICc: 4.84, BIC: 4.94, log likelihood = − 228.45
  3. B) AIC: 4.34, AICc: 4.34, BIC: 4.42, log likelihood = − 201.34
  4. C) AIC: 4.43, AICc: 4.46, BIC: 4.71, log likelihood = − 197.43