All reputable PV design models take temperature into account. There are 2 free such tools from NREL: PVWatts (simple) and SAM (detailed). RETScreen from Canada is also free. Some commercial tools are more sophisticated: Archelios, PVSyst, etc.
Soiling is more a site-specific parameter. Research is only starting on this topic, so there is no comprehensive "geographic" model for that yet. Reputable PV design tools typically assign either a default (hidden) value for that or a user-defined value (in other words, the user needs to do an educated guess).
I don't think precip is an input to any PV model, although it normally affects soiling.
There are many software tools for sizing of PV system, Among them, very few are available as open source (PV Watts, SAM, RETScreen, PVGIS etc). Some of these open source tools give only basic analysis (Power, Energy, Capacity Factor, PR's, net payment. These results are analyzed by considering the temperature coefficient) and few give details analysis with respect to various losses in the system (Arrow Diagram) and detailed economic analysis.
Commercial tools like HOMER (Hybrid combination of energy system can be simulated), PVSyst, PVSOL are very popular (in this tools 3-D modeling of the PV System is also possible.
All these have their own geographical data sets which consist of many parameters (Solar radiation, sunshine hours, temperature, wind speed, humidity, cloud cover etc). But I am not sure whether rainfall data is considered or not....?
But I am not sure whether rainfall data is considered or not....?
Soiling effect is considered in the commercial tools like PVSyst and PVSOL