Spatiotemporally consistent global dataset of the GIMMS Leaf Area Index (GIMMS LAI4g) from 1982 to 2020 (V1.2)

Brief Introduction:

The fourth generation GIMMS Leaf Area Index product (GIMMS LAI4g, version 1.2) provides spatiotemporally consistent global LAI data in half-month and 1/12° from 1982 to 2020. It is created to address two major uncertainties presented in current global long-term LAI products, i.e., (1) the effects of NOAA satellite orbital drift and AVHRR sensor degradation and (2) insufficient LAI reference data to build robust LAI model particularly before the late 1990s.

The GIMMS LAI4g was generated based on biome-specific BPNN models that employed the latest PKU GIMMS NDVI product and 3.6 million high-quality global Landsat LAI samples. It was then consolidated with the Reprocess MODIS LAI to extend the temporal coverage to 2020 via a pixel-wise Random Forests fusion method.

The GIMMS LAI4g exhibits overall high accuracy and low underestimation evaluated by field LAI measurements and Landsat LAI samples. It efficiently eliminated the effects of satellite orbital drift and sensor degradation and presents a good temporal consistency before and after the year 2000 and a more reasonable global vegetation trend. It could potentially facilitate mitigating the disagreements between studies of the long-term global vegetation changes and benefit the model development in Earth and environmental sciences.

Here we provide two versions of GIMMS LAI4g for download, one solely based on AVHRR data (1982−2015) and the other consolidated with the Reprocess MODIS LAI (1982−2020). We strongly recommend an adequate use of the quality control (QC) layer in the product. Please refer to the Readme file for more details.

Major updates:

Version 1.0 (February 17, 2023):

· The original version of the product.

Version 1.1 (June 14, 2023):

· The GIMMS LAI4g is now validated by ground LAI measurements.

· A pixel-wise Random Forests consolidation method is used to replace the linear one.

· Two versions of GIMMS LAI4g are now available, one solely based on AVHRR data and one consolidated with MODIS LAI.

Version 1.2 (August 25, 2023):

· The BPNN model without explanatory variables of NOAA satellite number and years since launch is used to generate LAI values during 1982−1984 for all biomes, October−April for EBF, and winters for ENF, when the Landsat NDVI samples were absent or relatively scarce.

Dataset Characteristics:

Spatial Coverage:       180ºW~180ºE, 63ºS~90ºN

Projection:                  Geographic

Spatial Resolution:     1/12 degree

Temporal Resolution: Half month

Temporal Coverage:   January 1982 to December 2020

Image Dimension:      Rows-2160; Columns-4320

Units:                          m²/m²

Fill Value:                   65535

Data Type:                  uint16

Valid Range:              0-7000

Scale Factor:                     0.001

File Format:                TIFF(.tif)

File Size:                    ~8Mb each file

References:

Cao, S., Li, M., Zhu, Z., Wang, Z., Zha, J., Zhao, W., Duanmu, Z., Chen, J., Zheng, Y., Chen, Y., Myneni, R. B., and Piao, S.: Spatiotemporally consistent global dataset of the GIMMS Leaf Area Index (GIMMS LAI4g) from 1982 to 2020, Earth Syst. Sci. Data Discuss. [preprint], https://doi.org/10.5194/essd-2023-68, in review, 2023.