Preparation Of AVHRR Snow Cover Extent Dataset In The Northern Hemisphere And Characteristics Of Snow Phenology

Snow is an important component of the cryosphere and one of the important factors affecting the atmospheric environment.Snow cover extent is one of the important snow cover parameters,which is an important input parameter of many global climate models.Due to the lack of snow cover extent dataset with slightly high spatio-temporal resolution and high accuracy in the Northern Hemisphere from 1988 to 1999,we used NOAA-AVHRR CDR version 4 surface reflectance data as the basic input data and Landsat-5 TM snow maps as reference maps to obtain the optimal threshold of snow discriminant algorithm based on multilayer decision tree,establishing a snow algorithm suitable for the northern hemisphere.In addition,due to the complete lack of AVHRR dataset in high latitudes and partial lack of data in low latitudes,we filled the gaps of AVHRR L1 dataset by the Northern Hemisphere0.25° snow depth dataset,combination of Hidden Markov Model de-clouding,temperature and water mask were used to generate the Northern Hemisphere AVHRR 1988—1999 5 km daily snow cover extent dataset.The accuracy of the dataset is evaluated by using meteorological stations and high-resolution remote sensing images.Finally,the snow cover phenological parameters are calculated and the spatio-temporal variation is analyzed.The main research work and results are as follows:(1)Production of Northern Hemisphere snow extent dataset.Taking Landsat-5 TM image as a sample,The optimal threshold of the snow identification algorithm was obtained for the Northern Hemisphere,establishing a snow algorithm suitable for the northern hemisphere,and we also used the cloud discrimination algorithm,and then produced the Northern Hemisphere AVHRR L1 daily snow cover extent dataset.Then,we filled the gaps of AVHRR L1 dataset by the Northern Hemisphere 0.25° snow depth dataset and generated the Northern Hemisphere AVHRR L2 daily snow cover extent dataset.Finally,the Hidden Markov Model was used to construct an energy-time function for de-clouding,and the temperature mask and water mask were combined to generate the L3 daily snow cover extent dataset,i.e.,the AVHRR daily cloud-free snow cover extent dataset for the Northern Hemisphere from 1988 to 1999.(2)Accuracy assessment of snow cover extent datasets in the northern hemisphere.Firstly,the accuracy of the AVHRR snow extent dataset at all levels was evaluated using snow depth data from meteorological stations,and the overall accuracy of the dataset at all levels was high,with the overall accuracies of 83.73%,86.12% and 87.20% at L1,L2 and L3 daily snow cover extent dataset,respectively.Secondly,the accuracy of the datasets at all levels was verified by TM reference snow range images,and the overall accuracy was 82.50%,86.24% and 89.48%,respectively,with high overall verification results.In addition,for the AVHRR L3 daily snow cover extent dataset,the site validation accuracy and TM validation accuracy of typical snow cover areas in the Northern Hemisphere were further analyzed,and the overall accuracy was 84.32% and 89.92%,respectively.Finally,the accuracy site verification accuracy and TM verification accuracy of different surface types were analyzed.The results show that the AVHRR snow cover extent dataset basically matches with the reference true value,and the ability of snow identification is stronger under the coverage of cropland and grassland type,while there is a certain omission error under the forest coverage,but the overall accuracy of the northern hemisphere snow cover extent dataset is high and the dataset has certain reliability.(3)Analysis of spatial and temporal variability of snow phenology in the Northern Hemisphere Phenological.The daily,monthly and annual trends of snow cover area,and the spatial and temporal variation patterns of the start of snow cover days,the melt of snow cover days and snow cover days were calculated to reveal the snow cover distribution and change characteristics in the northern hemisphere.The results show that the annual average snow area in the northern hemisphere is slightly decreasing,and the annual average snow area,the annual average snow area maximum,and the annual average snow area minimum have the same overall trend and show certain seasonal patterns;Influenced by the global climate and other factors,the start of snow cover days,the melt of snow cover days and snow cover days all show certain latitudinal characteristics,among them,snow cover days increases overall with increasing latitude except for high mountainous areas,with a slightly decreasing trend in the middle latitude and a slightly increasing trend in the high latitude.There are obvious regional differences between the start of snow cover days and the melt of snow cover days,with the overall characteristics that the higher the latitude and the higher the altitude,the earlier the start of snow cover days,the later the melt of snow cover days,and the longer snow cover days.