Corn has become one of the most important crops in China at present.The study area in this work,which includes Jilin Province,Liaoning Province,Heilongjiang Province,and the four eastern cities of Inner Mongolia Autonomous Region(Hulun Buir,Tongliao,Chifeng,and Xing'an League),is the most important corn production area.The area is also located at the northern limit of corn planting area.Notably,the temporal and spatial distribution of chilling damage is highly important to increase yield and quality.This study aims to integrate MODIS and meteorological data for monitoring corn chilling damage in Northeast China.The algorithm was computed in two steps.In the first step,the remote sensing estimation model of air temperature were established.In the second step,the sterile-type chilling damage and delayed-type chilling damage on corn were monitored based on the full coverage daily mean air temperature and the corn chilling damage indicator.Satellite data,including LST,EVI,and quality control data derived from TERRA/AQUA-MODIS,and ground-based data,including daily mean air temperature and phenological data observed by 234 meteorological stations,from 2003 to 2015 were collected for data analysis,image processing,and mapping.The remote sensing estimation model of air temperature was established by multi-variated linear regression using the MODIS LST,EVI,and solar declination of cloud-free pixels as independent variables,and daily mean air temperature observed by meteorological stations was used as a dependent variable.The meteorological stations were divided into two parts according to the coordinates.Daily mean air temperature measured by two thirds of station(156)from 2003 to 2013 was used to establish the daily average temperature estimation model,and the remaining data including the observations of 78 meteorological stations from 2003 to 2013 and the observation data of all stations in 2014 and 2015 were used to validate the model.The MODIS EVI production is the composited production on the 16th day.The S-G filter with max was used to achieve daily EVI.The air temperature of cloud-free pixels was calculated with estimated models using TERRA and AQUA daytime and nighttime data,respectively.Then,the TERRA and AQUA daytime and nighttime derived daily mean air temperature data were merged based on the R2 and RMSE to increase spatial cloud-free data.The validation results show that the models using TERRA or AQUA night LST data as predictors outperform those using daytime LST as predictors.A framework was proposed for the air temperature data fusion.The data fusion framework is based on the fact that the MODIS TERRA and AQUA can provide daytime and nighttime LST and that the merge of these data can increase spatial coverage.The daily mean air temperature dataset covering the whole study area with a spatial resolution of 1 km from 2003 to 2015 was completed based on the retrieval models and the data fusion framework.This study provided a remote sensing monitoring method of sterile-type chilling damage and delayed-type chilling damage on com.Low daily mean air temperature and its last days are the corn sterile-type chilling damage indicator.Corn sterile-type chilling damage was identified by integrating the corn chilling damage indicator and the daily mean air temperature dataset.The results showed corn sterile-type chilling damage in 2003,2006,and 2012.These findings are consistent with the meteorological observation.This research can be used to monitor the process of com sterile-type chilling damage,take abatement measures to mitigate corn sterile-type chilling damage,and reduce disaster losses.On the basis of the full coverage daily mean air temperature dataset,the accumulated temperature of≥10 ℃ from 2003 to 2015 was calculated.The indicators of delayed-type chilling damage on corn also revealed that the study area suffered from widespread delayed-type chilling damage in 2003,2005,2006,2009,and 2011.Compared with the observation of meteorological stations,the results of this research match the actual situation.
维普
万方数据
