The spectroscopic properties of water vapour are of crucial importance for the understanding of the energy balance in the Earth atmosphere because water is the most important absorber of both incoming and outgoing radiation. However models of atmospheric absorption using available spectroscopic data significantly underestimated the total absorption. Water has many, many weak transitions, which are routinely neglected and it is possible that these are responsible for much of the missing absorption. We will evaluate the contribution of weak water vapour lines to the absorption of solar radiation in the Earth's atmosphere and determine their contribution in different spectral intervals, especially in the atmospheric transparency windows. A new detailed database o f water transitions will be created using the latest high quality potential energy and dipole moment surfaces constructed at the host Institute. Precise, synthetic line-lists calculated using variational methods, together with calculations performed using a n effective Hamiltonian approach will be used to extend the spectroscopic information about H2O absorptions, particularly by weak lines, over a wide spectral region. In particular, spectroscopic data on isotopically substituted water molecules, such as H21 8O, H217O and HDO, will be calculated over the entire visible region. This resulting improved database of water transitions will be used in calculations of atmospheric absorption of solar radiation.These calculations will be performed in the wide spectral region from 10000 to 20000cm-1 and will use a realistic atmospheric model, which includes diffuse light scattering in a cloudless atmosphere. The final result will be the first realistic calculation of atmospheric absorption with accuracy better 1% for a clear and cloudless atmosphere. This result will be tested against direct observations of atmospheric spectra obtained both from the ground and via remote sensing satellites.