Abstract:
Using eddy covariance data for May to October 2005, 2006 and 2007 collected from rice paddies in the Sanjiang Plain, diurnal and seasonal variations in energy flux, partitioning and balance closure were analyzed. The results show that diurnal varia-tions in net radiation and latent heat flux are uni-peak model. However, a relatively distinct diurnal variation in sensible heat flux isobserved only at the maturity stage of rice. There is no obvious diurnal variation in soil heat flux in the entire growing season. Seasonal variations in net radiation are significant, generally reaching the maximum of 18~20 MJ·m
-2·d
-1 during late June to early July. Seasonal variations in latent heat flux are similar to those in net radiation, with a maximum range of 13~19 MJ·m
-2·d
-1. In contrast, sensible heat flux is smaller and less variable on seasonal scale (-3.90~3.94 MJ·m
-2·d
-1) than latent heat flux. Soil heat flux is not only small, but has a declining trend for May to October, and with a range of -2.67~3.62 MJ·m
-2·d
-1. There is a sig-nificant energy partition in the rice paddies of the Sanjiang Plain, with mean
LE/
Rn ratio of 0.67 (for May to October). This suggests that most of the energy is latent heat flux, but with a larger
LE/
Rn ratio at blooming stage than at other growth stages. The seasonalpattern of
Hs/
Rn ratio is inversely related to
LE/
Rn ratio, with an average
Hs/
Rn ratio of 0.10. This implies that the Bowen ratio is smaller for blooming stage and larger for the early and late stages of growth.
G/
Rn ratio is small, showing a declining trend of 0.14~ -0.08 from May to October. The energy balance closure, evaluated by methods of Ordinary Linear Squares regression and Energy Balance Ratio, is generally small in the study area. Averaged energy imbalance ratio in the paddy fields for May to October isrespectively 22% and 16% for 2005 and 2006. However, energy balance ratio surprisingly hits 1.07 in 2007, showing that the sum of latent and sensible heat fluxes is larger than the available energy