The 2015 1km / 5-day synthetic leaf area index (LAI) data set of Heihe River basin provides the 5-day Lai synthesis results in 2015. The data uses Terra / MODIS, Aqua / MODIS, domestic satellites fy3a / MERSI and fy3b / MERSI sensor data to build a multi-source remote sensing data set with spatial resolution of 1km and temporal resolution of 5 days
Multi-source remote sensing data sets can provide more angles and more observations than a single sensor in a limited time. However, the observation quality of multi-source data sets is uneven due to the difference of on orbit running time and performance of sensors. Therefore, in order to make more effective use of multi-source data sets, the algorithm first classifies the quality of multi-source data sets, which are divided into primary data, secondary data and tertiary data according to the observation rationality.
Level III data are observations polluted by thin clouds and are not used for calculation. The purpose of quality evaluation and classification is to provide basis for the selection of optimal data set and the design of inversion algorithm flow in Lai inversion. The inversion algorithm of leaf area index product is designed to distinguish mountain and flat land and vegetation types, and the neural network method of different models is used for inversion.
Based on global DEM map and surface classification map, PROSAIL model is adopted for continuous vegetation such as grassland and crops, and slope gost model is adopted for forest and mountain vegetation. The reference map generated by using the ground measured data of forests in the upper reaches of Heihe River and oases in the middle reaches of Heihe River, and the corresponding high-resolution reference map is scaled up to 1km resolution. Compared with Lai products, the products have good correlation between farmland and forest areas and the reference value, and the overall accuracy basically meets the accuracy threshold that the error specified by GCOS does not exceed (0.5, 20%). This product is cross compared with Lai products such as MODIS, geov1 and glass. Compared with the reference value, the accuracy of this Lai product is better than that of similar products.
In short, the 1km / 5-day synthetic Lai data set in Heihe River basin makes comprehensive use of multi-source remote sensing data to improve the estimation accuracy and time resolution of Lai parameter products and better serve the application of remote sensing data products
The 2015 1km / 5-day synthetic leaf area index (LAI) data set of Heihe River basin provides the 5-day Lai synthesis results in 2015. The data uses Terra / MODIS, Aqua / MODIS, domestic satellites fy3a / MERSI and fy3b / MERSI sensor data to build a multi-source remote sensing data set with spatial resolution of 1km and temporal resolution of 5 days
The inversion algorithm of leaf area index product is designed to distinguish mountain and flat land and vegetation types, and the neural network method of different models is used for inversion
Based on global DEM map and surface classification map, PROSAIL model is adopted for continuous vegetation such as grassland and crops, and slope gost model is adopted for forest and mountain vegetation. The reference map generated by using the ground measured data of forests in the upper reaches of Heihe River and oases in the middle reaches of Heihe River, and the corresponding high-resolution reference map is scaled up to 1km resolution. Compared with Lai products, the products have good correlation between farmland and forest areas and the reference value, and the overall accuracy basically meets the accuracy threshold specified by GCOS that the error does not exceed (0.5, 20%)
Good data quality
| # | number | name | type |
| 1 | 2013AA12A301 | National High-tech R&D Program of China (863 Program) | |
| 2 | 2012AA12A304 | National Natural Science Foundation of China |
This work is licensed under a
Creative
Commons Attribution 4.0 International License.
CSTR:11738.11.ncdc.NIEER.2021.37
10.12072/ncdc.nieer.db3553.2023
