Abstract: Abstract: The decrease of solar radiation and increase of diffuse radiation proportion caused by haze affect crop productivity. To investigate the impacts of changed radiation by haze on the crop growth, a device that simulating haze-caused radiation changes was designed and tested in an open field during 2013 and 2014. Based on air quality index (AQI), haze is classified into three levels: mild, moderate and heavy haze. The changes of solar radiation under each level of haze condition are determined based on the relationship between AQI and radiation conditions. Global radiation decreases by 11%-21%, 22%- 32%, 33%-54% under mild, moderate, heavy haze while corresponding diffuse radiation proportion are 51%-59%, 60%-68%, 69%-87%, respectively. The optical properties (transmittance, scattering and spatial distribution of visible spectrum) of different thickness and layers of PE film used in the device were experimentally tested as covering materials to simulate radiation conditions under the moderate and heavy haze conditions. Based on the solar height and azimuth angles during crop growing season, the minimum plot area was determined as 20 m2 to secure 4 m2 shading area in a plot from 10:00 to 14:00. The haze simulator was designed as a cuboid with 5 m (east-west)×4 m (north-south) × 2 m (above the ground) with the cover material installed on top and without obstacle at the 4 sides to allow free air flow and heat exchange. The device was applied and tested in rice paddy field during two rice growing seasons (2013 and 2014). Solar radiation and other microclimate factors were automatically monitored and 30 minutes average values were recorded during the experiments. The results showed that the changes of total radiation and diffuse radiation proportion under different treatments matched the change scopes of solar radiation under moderate and heavy levels of haze. The spatial distribution of visible spectrum under treatment conditions was not significantly different from that under natural haze conditions. The differences of air, water and soil temperature between shading treatments and control were less than ±0.5℃ in most of the time, and the differences of air humidity between shading treatments and control were 4%-5%. The canopy thermal image showed no significant difference between shading treatment and control at rice heading stage in 2014. These results indicated that the device designed in this study can be used to simulate haze-caused radiation conditions in open field and has the potential to provide effective field experimental means for investigating the effects of haze on crops.