In response to the accident at Tokyo Electric Power Company's Fukushima Dai-ichi Nuclear Power Plant (FDNPP), vehicle-borne monitoring was used to map radiation levels for radiological protection of the public. By convention measurements from vehicle-borne surveys are converted to the ambient dose equivalent rate at 1 m height in the absence of the vehicle. This allows for comparison with results from other types of survey, including surveys with hand-held or airborne instruments. To improve the accuracy of the converted results from vehicle-borne surveys, we investigated combining measurements from two detectors mounted on the vehicle at different heights above the ground. A dual-detector setup was added to a JAEA monitoring car and compared against hand-held survey meter measurements in Fukushima Prefecture. The results obtained by combining measurements from two detectors were within ±20% of the hand-held reference measurements. The mean absolute percentage deviation from the reference measurements was 7.2%. The combined results from the two detectors were more accurate than those from either the roof-mounted detector, or the detector inside the vehicle, taken alone. One issue with vehicle-borne surveys is that ambient dose equivalent rates above roads are not necessarily representative of adjacent areas. This is because radiocesium is often deficient on asphalt surfaces, as it is easily scrubbed off by rain, wind and vehicle tires. To tackle this issue, we investigated mounting heights for vehicle-borne detectors using Monte Carlo gamma-ray simulations. When radiocesium is deficient on a road compared to the adjacent land, mounting detectors high on vehicles yields results closer to the values adjacent to the road. The ratio of ambient dose equivalent rates reported by detectors mounted at different heights in a dual-detector setup indicates whether radiocesium is deficient on the road compared to the adjacent land.