Photovoltaic (PV) system installation on freshwater bodies like canals, instead of on-land installation, can help preserve valuable land. Canal site microclimate can induce localized cooling, which improves PV performance due to a reduced solar cell temperature (T-cell). This impact has been studied in the literature using mostly ex-situ experimental setups that approximate the canal site microclimate using water tanks. This work presents an alternative approach where a robust microclimate prediction model is developed based on standalone measurements in a hot semi-arid climate. This model evaluates canal-site ambient conditions for different regions, climates, and months characterized by the land-site ambient temperature (T-amb) and relative humidity (RH) levels between 13.5-46.3 degrees C and 12%-76% respectively. The corresponding land-site specific humidity (SH) range is 4.95-15.73 g/kg. This prediction model is later used with an optimized three-dimensional distributed steady-state (3DDSS) PV model to analyze the annual performance of canal- and land-site PV using thermo-electric simulations. With a single 670 W commercial PV panel, an annual increase in 1.52% in the electrical energy generation with a T-cell reduction of 4.52 degrees C is obtained for the Rakh Branch canal site, Pakistan. In situ experimental measurements have also been performed with a scaled-down PV setup, which supports these results. This work provides mathematical models for researchers to use in conjunction with public ambient condition databases to estimate microclimate conditions over small freshwater bodies. It also acts as a guideline for consumers and public policymakers to evaluate the potential of canal site PV systems accurately.