Skip to main content
. 2024 Jan 18;68(4):701–717. doi: 10.1007/s00484-023-02616-y

Table 3.

Hidden Markov Model results for regime change in spring climatology at the Cloquet Forestry Center (CFC) and Marcell Experimental Forest (MEF) for both station and Climate Research Unit (CRU) climate data. Model performance for the one-state and two-state model was evaluated by the change in Akaike Information Criterion (Δ AIC). Model log-likelihood is also indicated for all models and is compared among models using a log-likelihood test ( = p < 0.1, *p < 0.05, **p < 0.01). The transition year (T) is indicated if a single transition year was identified. The superior model is in bold

Record n state = 1 n state = 2 Model selection
AIC BIC Log L AIC BIC Log L Δ AIC T Year
CFC station 385.7 390.97 −190.85 374.07 392.52 −180.04** −11.63 1976
MEF station 221.62 225.6 −108.81 224.63 238.56 −105.32 3.01
CFC CRU 438.54 444.01 −217.27 423.06 442.21 −204.53** −15.48 1976
MEF CRU 452.89 458.36 −224.44 444.66 463.81 −215.33** −8.23 1976
Models with CMIP6 + PNA covariates (1961–2014)
Record n state = 2 n state = 2 + covariates
AIC BIC Log L AIC BIC Log L Δ AIC T Year
CFC station 259.45 274.67 −122.72 258.17 277.74 −120.08 • −1.28 1976
MEF station 224.63 238.56 −105.32 226.42 244.32 −104.21 1.79
CFC CRU 261.97 277.19 −123.99 260.45 280.02 −121.22 • −1.52
MEF CRU 273.95 289.17 −129.98 273.32 292.89 −127.66 • −0.63