Global Navigation Satellite Systems (GNSS) provide positioning service to civil applications. At the end user's side, these positioning services will have different position accuracy levels when satellite signal configurations and receiver noise levels are different. In this paper, a position error modeling is presented using a black-box approach. Using this approach, end users will treat the entire signal biases and receiver noises as a black-box transfer function. No efforts are made to model many different individual error sources of satellite signals and receivers. Further, a different type of sensor is chosen to aid the error de-correlation. Experimental results in different configurations are presented.