Fig. 1
Study area map with locations of the observation stations (LB and LBS) and the photographs of the stations.
Fig. 2
Observed half-hourly T a (grey line), T w (red line) and T s (green line, as calculated from the upward longwave radiation) at LS in (a) 2011, (b) 2012 and (c) 2013.
Fig. 3
(a) Observed monthly mean diurnal cycles of temperature differences between lake skin surface and air ΔT sa; (b) 15-d mean temperature differences (bars) and standard deviations (lines) between the lake surface and air ΔT sa in 2011–2013; (c) Box–Whiskers plot of diurnal Ta amplitudes T 24 in the 3 years. The tops and bottoms of each ‘box’ are the 25th and 75th percentiles, respectively. The line in the middle of each box is the median. The whisker lengths correspond to±3 times standard deviation assuming normal distribution of the data with outliers marked as points; (d) diurnal course of air and lake surface temperatures on 12 June 2013 (the day with maximum diurnal T a amplitude in 2013); (e) diurnal course of T a and T s on 16 May 2013 (the day with minimum diurnal T a amplitude in 2013); (f) the cool-skin temperature drop for the day from (d) and (e).
Fig. 4
(a) Observed (black solid line) and simulated (colour lines) T s, (b) T w, (c) ΔT sa, (d) ice thickness, (e) H, and (f) LE in CTL (blue dash line) and sensitivity experiments Sd2 (red dotted line) and Sd32 (green line) with different lake depths.
Table 1. Bias, RMSE and cc between simulation and observation
BiasSd22.2 °C1.2 °C14.2Wm−238.9Wm−20.14mCTL1.4 °C0.4 °C9.8Wm−231.2Wm−2−0.05mSd321.2 °C0.2 °C8.7Wm−229.1Wm−2−0.21mRMSESd22.6 °C1.9 °C17.3Wm−243.2Wm−20.15mCTL1.6 °C0.7 °C14.7Wm−235.2Wm−20.07mSd321.5 °C0.6 °C14.1Wm−233.9Wm−20.21mccSd20.880.830.630.680.91CTL0.950.980.700.860.97Sd320.940.980.700.850.91
Fig. 5
(a) RMSE (dashed line) and cc (solid line) between the simulated and observed Ts in the sensitivity experiments with Lec changing from 0.1m−1 to 1.2m−1 and in CTL (Cross), (b) observed (black solid line) and simulated T s, and (c) ΔT sa in CTL (blue solid line) and sensitivity experiments Slec0.1 (red dashed line) and Slec 1.2 (green dotted line) with 0.1m−1 and 1.2m−1 light extinction coefficient.
Fig. 6
(a) Simulated T s in CTL (blue solid line) and the sensitivity experiment Sta (red dashed line) and simulated T s difference (brown dotted line) between CTL and Sta, (b) simulated ΔT sa in CTL (blue solid line) and Sta (red dashed line), and (c) simulated H (blue solid line) and LE (red dashed line) difference between CTL and Sta.
Fig. 7
Same as Fig. 6, but for CTL and the sensitivity experiment Sswd.
Fig. 8
Same as Fig. 6, but for CTL and the sensitivity experiment Slwd.
Fig. 9
(a) Simulated H (blue solid line) and LE (red dashed line) difference between CTL and the sensitivity experiment Sq, and (b) simulated ΔT sa in CTL (blue solid line) and Sq (red dashed line).
Fig. 10
Same as Fig. 9, but for CTL and the sensitivity experiment Sps.
Fig. 11
Changed percentage of annual mean ΔT sa from July 2012 to June 2013 in each simulation experiment (with percentage change of each variable) compared to that of CTL.