With the support of the first topic "sharing and integration of three pole big data" (xda19070100) of the special space-time three pole environment project of the earth big data science project, Che Tao research group of Northwest Institute of ecological environment and resources, Chinese Academy of Sciences uses machine learning methods combined with multi-source snow depth product data The daily snow depth data set of long-time Series in the northern hemisphere is prepared. Firstly, the applicability of artificial neural network, support vector machine and random forest method in snow depth fusion is compared. It is found that random forest method has strong advantages in snow depth data fusion. Secondly, using the random forest method, combined with remote sensing snow depth products such as AMSR-E, amsr2, NHsd and globsnow and reanalysis data such as era interim and merra2, the grid snow depth products and environmental factor variables are used as the input independent variables of the model, and the data of China Meteorological Station (945), Russia meteorological station (620) and Russia snow survey data (514) The snow depth data of 43340 ground observation stations such as the daily data of the global historical meteorological network (41261) are used as the reference truth to train and verify the model, and the daily grid snow depth data set of the snow hydrological year from 1980 to 2019 (September 1 of the previous year to May 31 of the current year) is prepared on the cloud platform provided by the special "earth big data science project". Using the global snow model comparison program and independent ground observation data for verification, the quality of the fusion data set has been improved as a whole. According to the comparison between the ground observation data and the snow depth products before fusion, the determination coefficient (R2) of the fusion data is increased from 0.15 (globsnow snow depth products) to 0.91, and the corresponding root mean square error (RMSE) and mean absolute error (MAE) are also reduced to 5.5 cm and 2.2 cm. The following is the header file content of each file. Adding it to the front of each file can display the data in ArcMap. Ncols 1440 / / 1440 columns in the data matrix Nrows 360 / / the data matrix has 360 rows in total Xllcenter - 180 / / the corner coordinates of the grid at the lower left corner of the matrix in the X direction Yllcenter 0 / / the corner coordinates of the grid at the lower left corner of the y-direction axis of the matrix Cellsize 0.25 / / size of each grid NODATA_ Value - 9999 / / default value
CHE Tao, HU Yanxing, DAI Liyun, XIAO Lin
This biophysical permafrost zonation map was produced using a rule-based GIS model that integrated a new permafrost extent, climate conditions, vegetation structure, soil and topographic conditions, as well as a yedoma map. Different from the previous maps, permafrost in this map is classified into five types: climate-driven, climate-driven/ecosystem-modified, climate-driven/ecosystem protected, ecosystem-driven, and ecosystem-protected. Excluding glaciers and lakes, the areas of these five types in the Northern Hemisphere are 3.66×106 km2, 8.06×106 km2, 0.62×106 km2, 5.79×106 km2, and 1.63×106 km2, respectively. 81% of the permafrost regions in the Northern Hemisphere are modified, driven, or protected by ecosystems, indicating the dominant role of ecosystems in permafrost stability in the Northern Hemisphere. Permafrost driven solely by climate occupies 19% of permafrost regions, mainly in High Arctic and high mountains areas, such as the Qinghai-Tibet Plateau.
RAN Youhua, M. Torre Jorgenson, LI Xin, JIN Huijun, WU Tonghua, Li Ren, CHENG Guodong
In order to understand the temporal and spatial variation characteristics of temperature changes in the Northern Hemisphere, the study used CRU (Climatic Research Unit) grid data to calculate the spatial distribution of the average annual temperature of 30 years (1971-2000). The annual average temperature decreases with the latitude increasing, and varies from greater than 30 °C to less than -25 °C. In the regions of the same latitudes, the annual average temperature in high altitude areas (such as the Tibetan Plateau, the Mongolian Plateau, and the Western Siberian Mountains) presented the trend of low temperature. At the same time, the annual average temperature trend distribution map of the Northern Hemisphere with a resolution of 0.5 ° × 0.5 ° from 1901 to 2016 was completed.
YIN Guoan, SHI Yaya
Lake ice phenology is a seasonal cyclical feature that describes lake ice coverage. The change of lake ice phenology is an important part of carbon, water and energy process study, and one of the sensitive factors of climate change. This dataset is a lake ice phenology based on passive microwave inversion, including lake ice phenology of 200 lakes in the Tibetan Plateau and high latitudes area of the Northern Hemisphere from 2002 to 2018 (including freeze-up start date, freeze-up end date, break-up start date, and break-up end date of the lakes), data of some lakes can date back to 1978. This data is basically consistent with the MODIS monitoring results from the same time with an interpretation error of 2-4 days. Users can use this data to conduct climate change study in the Northern Hemisphere.
This dataset is the spatial distribution map of the marshes in the source area of the Yellow River near the Zaling Lake-Eling Lake, covering an area of about 21,000 square kilometers. The data set is classified by the Landsat 8 image through an expert decision tree and corrected by manual visual interpretation. The spatial resolution of the image is 30m, using the WGS 1984 UTM projected coordinate system, and the data format is grid format. The image is divided into five types of land, the land type 1 is “water body”, the land type 2 is “high-cover vegetation”, the land type 3 is “naked land”, and the land type 4 is “low-cover vegetation”, and the land type 5 is For "marsh", low-coverage vegetation and high-coverage vegetation are distinguished by vegetation coverage. The threshold is 0.1 to 0.4 for low-cover vegetation and 0.4 to 1 for high-cover vegetation.
HUANG Xiaodong, DAI Liyun