Abstract—This study performs spatial prediction of air quality monitoring data. In addition to CO2, which is the target of prediction, air quality monitoring sensors measure temperature, humidity, illuminance, atmospheric pressure, and so on. Using data acquired from multiple air quality sensors and distributed compression sensing, we will predict changes in carbon dioxide concentration at unmeasured points (hypothetical), and compare the results with actual measurements to evaluate the accuracy. If the prediction is accurate, it is expected that the number of sensor nodes to be deployed can be reduced, leading to energy savings.