Far field projections
The near to far field projections calculate the EM fields anywhere in the far field. The near field data is obtained from one of Lumerical's optical solvers. The Far Field Projections (FFP) are a postprocessing step to calculate the electromagnetic fields at points in space that lie far away from a structure generating light. A typical far field projection samples the fields near the radiating structure and propagates them to any requested point in space. The near fields are usually collected from Lumerical's optical solvers using frequency domain monitors. If your structure is periodic, consider using the Grating projections features instead.
A simple way to understand far field projections is to view them as a decomposition of the near field data using a set of plane waves propagating at different angles as the basis for the decomposition. The end result is that the far field projections functions provide a straightforward and numerically efficient method for calculating the EM fields anywhere in the intermediate and far field regions.
Near to far field projection application
The example below shows how the far field projection can be used to see the angular distribution of reflected light from a defective surface. The light is reflected off a "bump" on the surface from a focused beam that is slightly misaligned with respect to the surface normal. The simulation can show how the fields interact with the defect (bump). The reflected light can be recorded in the near field slightly above the surface. After the simulation is complete, the far field projection can be used to project the reflected radiation to a surface (a hemisphere in this example) at a far away distance. The final result is an image of the field intensity on the hemisphere, as viewed from above. The effects of the misalignment of the spot can clearly be seen from the far field projection results.
Understanding when far field projections can be used
Requirements:
 The EM fields must be known everywhere on a plane or on a closed surface.
 The plane or closed surface must be in a single homogeneous material.
 The material must extend out to infinity. There can be no additional structures (or sources) beyond this plane or closed surface.
 The far field projection functions assume that the EM fields are zero beyond the edge of the monitor. This effectively truncates the near fields at the monitor edge. See this page for spatial filtering technique (Currently available in FDTD and MODE only)
 The material can be dispersive as long as the loss is negligible, ie k<<n. However, the loss is not taken into account for the projection to the far field.
There are two types of situations where these conditions are satisfied:
Case 1  fields known on a plane 

Case 2  fields known on a closed surface 



When fields are known on a single surface, the projection functions can be used to calculate the fields anywhere beyond that surface. In the above screenshot, a monitor is located above the source. This monitor records all of the reflected fields. There are no additional structures or sources of light above the source. In this situation, the far field projections can be used to calculate the EM fields anywhere above the monitor plane. 

When the fields are known on a closed surface, the projection functions can be used to calculate the fields anywhere beyond that surface. In the above screenshot, a monitor surrounds the source and scattering particle. The monitor records all reflected fields. There are no additional structures or sources of light above the source. In this situation, the far field projections can be used to calculate the EM fields anywhere outside of the monitor object. 
Far field projection from multiple monitors
The screen shot below demonstrates the an example far field projection performed in FDTD from near field data known on a plane.
The monitors that are used for recording near field data for far field projection have a finite width, but in principle the far field projections need to know the fields everywhere on a plane extending to +/ infinity. The projection functions assume the fields are zero everywhere beyond the edge of the monitor. This assumption is only accurate when the monitor is wide enough to record most of the light that is propagating into the far field. Using a monitor that is too narrow will lead to errors in the far field data.
The truncation that might happen as a result the limit in the extent of the monitors might lead to high frequency ripple in the far field projection data. The far field filter option available in FDTD and MODE allows us to apply a spatial filter to the near field data. When the fields go more smoothly to zero, the high frequency ripple is removed from the projection. See the Far field filter page for more information.
The screen shot below demonstrates the an example far field projection performed in FDTD from near field data known on a closed surface. The far field projection data can be visualized as a crosssection on a specific plane in the form of a polar plot as shown in the figure below.
In the case of far field projection from fields known on a closed surface, a box of monitors is needed to record near field data in 3D space and the integration of all the data recorded by all the monitors in the box will be used for projection. See far field from a box analysis group for details. Only the farfieldexact, farfieldexact2d and farfieldexact3d commands allow projections from multiple monitors added to create a total far field projection.
Related publications
Allen Taflove, Computational Electromagnetics: The FiniteDifference TimeDomain Method. Boston: Artech House, (2005).
See also
Far field projection scripting commands