We explore the impact of agriculture, forest and cloud feedback on the surface energy budget using data obtained using a research aircraft, mesonet towers and model data. The forest has an order of magnitude larger roughness length, a lower albedo, a much smaller seasonal cycle in surface Bowen ratio (BR) and a weak mid-summer maximum of CO 2 uptake compared to agricultural areas, which have much smaller BR and much higher mid-summer CO 2 uptake, but a net CO 2 release and much reduced evaporation in spring and fall. Higher surface temperatures and the higher albedo over agricultural land reduce R net near local noon in the warm season by about 50Wm −2 in comparison with the adjacent boreal forest. The annual averaged R net , derived from 2 years of tower data, is 14Wm −2 less over grass sites than over forest sites. A reanalysis time-series for the BOREAS southern study area shows the coupling on daily timescales between the surface energy partition, the mean boundary layer depth, the cloud field and the long-wave and short-wave radiation fields. The albedo of the cloud field, the cloud short-wave forcing at the surface, varies over the range 0.1–0.8 with decreasing surface BR, and plays a major role in the surface energy budget. We estimate that this cloud feedback may increase albedo by 0.13 and reduce R net by 25Wm −2 in summer over agricultural land.