[1] Measurements of the energy balance components of a small boreal lake (area 0.041 km², mean depth 2.5 m) in southern Finland were performed during four open water periods (April–October) in 2005–2008. Turbulent fluxes of sensible and latent heat acquired using the eddy covariance technique were accompanied by net radiation and water heat storage measurements. In April the lake was near isothermal, whereas in May the development of a thermocline was enabled by dark water color and a sheltered location. The thermocline continued to deepen until September down to the depth of 3.5 m and prevented the deeper water from interacting with the atmosphere. The sensible heat flux was governed by the air‐water temperature difference and had its minimum in the afternoon (values down to −45 W m⁻²) and peaked in the early morning (values up to 32 W m⁻²). The monthly means ranged from −9 W m⁻² (April 2005) to 24 W m⁻² (July 2008). The diurnal variation of the latent heat flux, controlled by vapor pressure deficit, had an opposite diurnal phase with a maximum in the afternoon (values up to 120 W m⁻²) and minimum in the morning (values down to ∼2 W m⁻²). The monthly averages ranged from 4 W m⁻² (October 2008) to 77 W m⁻² (July 2006). Furthermore, the lake acted as a heat sink until July and August when the maximum heat content was about 230 MJ m⁻². The monthly energy balance closure varied from 57% to 112%, and the average closure was 82% (residual 16 W m⁻²) and 72% (residual 23 W m⁻²) for 2006 and 2007, respectively.

In this study the high-frequency loss of carbon dioxide (CO2) and water vapor (H2O) fluxes, measured by a closed-path eddy covariance system, were studied, and the related correction factors through the cospectral transfer function method were calculated. As already reported by other studies, it was found that the age of the sampling tube is a relevant factor to consider when estimating the spectral correction of water vapor fluxes. Moreover, a time-dependent relationship between the characteristic time constant (or response time) for water vapor and the ambient relative humidity was disclosed. Such dependence is negligible when the sampling tube is new, but it becomes important already when the tube is only 1 yr old and increases with the age of the tube. With a new sampling tube, the correction of water vapor flux measurements over a Scots pine forest in Hyytiälä, Finland, amounted on average to 7%. After 4 yr the correction increased strongly, ranging from 10%–15% during the summer to 30%–40% in wintertime, when the relative humidity is typically high. For this site the effective correction improved the long-term energy and water balance.Results suggest that the relative humidity effect on high-frequency loss of water vapor flux should be taken into account and that the effective transfer function should be estimated experimentally at least once per year. On the other hand, this high correction can be avoided by a correct choice and periodic maintenance of the eddy covariance system tube, for example, by cleaning or changing it at least once per year.

Long-term measurements of fluxes of sensible heat (H), latent heat (LE) and carbon dioxide (Fc) were made from December 2005 to August 2006 over an urban landscape in Helsinki, Finland using the direct micrometeorological eddy covariance technique. Three distinguished sectors of land-use cover (vegetation, roads and buildings) allowed comparisons of fluxes over different urban surfaces. The normalized standard deviation of wind and scalars as a function of atmospheric stability were typical for rough surfaces, as were turbulence spectra and cospectra. Footprint analysis was performed by a boundary-layer one and half-order closure model allowing for discrimination of surface and canopy sinks/sources and complex topography. Fluxes were analysed as average diurnal courses over winter, spring and summer periods. H exceeded LE reaching 300 W m-2 over urban and road surfaces in the summer and it was close to 100 W m-2 in the winter. LE was highest in the summer over vegetation cover attaining 150 W m-2. The emi...

[1] Significant amounts of terrestrial carbon are processed in lakes and emitted into the atmosphere as CO₂. However, due to lack of appropriate measurements the absolute role of lakes in the landscape as sinks or sources of CO₂ is still uncertain. We conducted the first long‐term, ecosystem‐level CO₂ flux measurements with eddy covariance technique in a boreal lake within a natural‐state catchment covering 5 years. The aim was to reveal the natural level of CO₂ flux between a lake and the atmosphere and its role in regional carbon cycling. On average, the lake emitted ca 10% of the terrestrial net ecosystem production of the surrounding old‐growth forest and the main immediate drivers behind the fluxes were physical rather than biological. Our results suggest that lakes are an integral part of terrestrial carbon cycling.

Long‐term measurements of sensible and latent heat and carbon dioxide fluxes were performed over a boreal lake in southern Finland using the direct micrometeorological eddy covariance (EC) technique. The water column was sampled weekly for dissolved carbon dioxide, and the CO₂ flux was estimated also applying the concentration gradient method. Temperature and oxygen profiles of the lake were measured twice a week. The measurements covered one full open‐water period from April to November 2003, and it is the longest continuous CO₂ record ever measured over a lake by EC. The sensible heat flux H was positive, that is, from the lake to the atmosphere, except in May, when it was >0 W/m² at night and <0 W/m² in daytime. The latent heat flux dominated clearly over H in spring and summer; that is, the Bowen ratio was less than 1. Higher‐moment turbulence statistics proved to be efficient in detection of frequent nonstationary situations. Applying the statistical criteria for CO₂ concentration and vertical wind speed, averaging over a 5‐min period and selecting only the wind direction with longest fetch, we could obtain lake‐representative CO₂ fluxes. Footprint analysis based on a closure model revealed that the source areas were relatively short because of the presence of turbulence generated by the surrounding forest, compared to a larger lake with an extended smooth surface. We observed a net CO₂ source of 0.2–0.4 μmol m⁻² s⁻¹ excluding July, when the flux was closer to zero. The results are consistent with the gradient method, based on more infrequent sampling, and both methods gave the same average flux, 0.2 μmol m⁻² s⁻¹, over the whole open‐water period.