For the power output per unit area at a certain location, you need to know things. The power per unit area per wavelength, this gives you the number of photons that hit the cell per unit area per wavelength, and you must know how many of these photons will result in an electron at the contact. The latter is called the external quantum efficieny spectrum (EQE) or incendent photon to charge efficiency (IPCE). The EQE depends on the type of cell you have. If you want to calculate the annual yield of a solar cell at a specific location, you also need to take into account the spectral changes of the solar spectrum during the day.
You can find a calcualtion tool at: http://re.jrc.ec.europa.eu/pvgis/apps4/pvest.php
I actually have used the method of PVGIS to obtain the PV potential for some locations. However, I need information on average output per unit area for PV modules. i.e. if a location has a potential for generating 1000kWh/kWp of electricity, what is the amount of electricity that can be generated per unit area (eg.) for a year?
When working with PV you have to be careful differenciating energy delivered by the PV panel "following laboratory conditions" (i.e., clean surface, nothing broken or damage, etc.) with the real energy delivered by the whole PV installation when subtracting the diverse electric losses and assuming imperfect conditions. In this draft you can find an estimation of the real power density of PV; if you have specific information about your installation and location you may adapt the factors and find your own power density estimation: http://www.eis.uva.es/energiasostenible/wp-content/uploads/2011/11/solar-energy-draft.pdf
What do you mean with unit area? If you have the result 1000kWh/kWp out of PVGIS you have the annual energy production per installed kWp. If you want to know more about the necessary area then you can estimate 8m²/kWp to 6m²/kWp for PV systems. The range depends on the used technology (thin film to high efficency modules). This is usable for a rough estimaton because you don´t have the information about the dirty, DC losses, inverter clipping etc.
All the above answers are correct, offering different levels of detail. Just one thing: if by unit area of PV installations you mean land area, as opposed to module area, you need to account for this. If you cover land with PV panels mounted flat on the ground, the areas are the same, but usually PV panels are mounted at an angle for optimum yield/module and then you need to leave gaps between modules so they don't shadow each other. I think a usual estimate is that you need 3x more land area than module area but probably more detailed calculators are available.
A good simple and free tool is PVWatts: http://pvwatts.nrel.gov/pvwatts.php
You normally have to enter the array size and obtain the energy produced for a specified location. Of course, this production is only a long-term average, so be careful, and consider the interannual variability if you must deliver a minimum annual quantity.