The rapid development and increase of low-cost sensors provide an unprecedented new way in understanding the space-time distribution of environmental quality. However, due to the relative low quality of observed data, there is an emergent need in investigating the potentials of the use of these highly uncertain data. In this talk, we proposed a framework in spatiotemporal environmental estimation with the use of uncertain data. This framework includes both data calibration for the uncertainty assessment and data fusion for the estimation. In some cases, the probabilistic distribution of data uncertainty can be complex and non-gaussian. In this case, we consider QMC method with a great potential to incorporate into the data fusion framework and to obtain the efficient evaluation of high-dimensional data uncertainties.