An algorithm for localization and navigation using a series of point measurements in a time-invariant non-uniform scalar field whose distribution is known in advance is presented in this research. For point measurements, various scalar physical quantities can be considered such as temperature, density, salinity, intensity of light or an electromagnetic field, etc. A spatial distribution of these physical quantities forms a field map which enables map-based navigation for a vehicle. The filter formulation for this field map-based navigation leads to a nonlinear estimation problem. Several nonlinear estimation techniques including the extended Kalman filter and the pariticle filter are considered in this research. In addition, the feasibility of using a Rao-Blackwellized particle filter is examined and its performance is demonstrated through numerical simulations in a synthetic scalar field.