Air pollution is a major factor affecting human life and living quality in arid and semiarid regions. This study was conducted in the Al-Ahsa district in the Eastern part of Saudi Arabia to measure carbon dioxide (CO2) concentration over different land-use types. Initially, the study's land use/land cover (LULC) was classified using the spectral characteristics of Landsat-8 data. Then, sensors were placed in five sites of different LULC types to detect CO2, air temperature, and relative humidity. The Friedman test was used to compare CO2 concentration among the five sites. Five LULC types were identified over the study area: date palm, cropland, bare land, urban land, and water. The results indicated that CO2 concentration showed a maximum mean value of 577 ppm recorded from a site dominated by urban lands. During the peak time of human transportation, a maximum value of 659 ppm was detected. The CO2 concentration mean values detected for the other LULC types showed 535, 515, and 484 ppm for the bare land, cropland, and date palm, respectively. This study's sensors and procedures helped provide information over relatively small areas. However, modelling CO2 fluctuations with time for LULC changes might improve management and sustainability.