Spatiotemporally consistent global dataset of the GIMMS Normalized Difference Vegetation Index (PKU GIMMS NDVI) from 1982 to 2020

Abstract

Abstract. Global products of remote sensing Normalized Difference Vegetation Index (NDVI) are critical to assessing the vegetation dynamic and its impacts and feedbacks on climate change from local to global scales. The previous versions of the Global Inventory Modelling and Mapping Studies (GIMMS) NDVI product derived from the Advanced Very High Resolution Radiometer (AVHRR) provide global biweekly NDVI data starting from the 1980s, being a reliable long-term NDVI time series that has been widely applied in Earth and environmental sciences. However, the GIMMS NDVI products have several limitations (e.g., orbital drift and sensor degradation) and cannot provide continuous data for the future. In this study, we presented a machine learning model that employed massive high-quality and global-wide Landsat NDVI samples and a data consolidation method to generate a new version of the GIMMS NDVI product, i.e., PKU GIMMS NDVI (1982−2020), based on AVHRR and Moderate-Resolution Imaging Spectroradiometer (MODIS) data. A total of 3.6 million Landsat NDVI samples that well spread across the globe were extracted for vegetation biomes in all seasons. The PKU GIMMS NDVI exhibits higher accuracy than its predecessor (GIMMS NDVI3g) in terms of R2 (0.975 over 0.942), mean absolute error (MAE: 0.033 over 0.074), and mean absolute percentage error (MAPE: 9 % over 20 %). Notably, PKU GIMMS NDVI effectively eliminates the evident orbital drift and sensor degradation effects in tropical areas. The consolidated PKU GIMMS NDVI has a high temporal consistency with MODIS NDVI in describing vegetation trends (R2 = 0.962, MAE = 0.032, and MAPE = 6.5 %). The PKU GIMMS NDVI product can potentially provide a more solid data basis for global change studies. The theoretical framework that employs Landsat data samples can facilitate the generation of remote sensing products for other land surface parameters.

Conflicting Changes of Vegetation Greenness Interannual Variability on Half of the Global Vegetated Surface

“…However, there have been recent efforts aimed at addressing issues in certain aspects of the algorithms (Jeong et al, 2023;Liu et al, 2017). Although all four long-term satellite remote sensing products claimed to use rigorous algorithms for reconstructing "high-quality" data globally, they also admitted their inability to obtain many reliable data in the tropics using data quality control flags (Didan et al, 2015;Li et al, 2023;Liu et al, 2012;Ma & Liang, 2022), mostly due to persistent cloud cover in the tropics. Based on this, we suspect that the lack of high-quality observations in the tropics could affect the robustness of satellite remote sensing observations.…”

“…We used six long-term and three short-term satellite remote sensing datasets in this study, covering four types of vegetation indicators -NDVI, LAI, SIF and VOD (Table 1). These remote sensing products include: (a) VIP15 NDVI product (1981, which has 0.05°and 15-day resolutions and is developed by harmonizing the observations of Advanced Very High Resolution Radiometer (AVHRR) from 1981 to 1999 and Moderate Resolution Imaging Spectroradiometer (MODIS) C5 from 2000 to 2014 (Didan et al, 2015); (b) GIMMS NDVI3g , which has 15-day and 1/12°resolutions, and was produced by aggregating daily AVHRR surface reflectance (Pinzon & Tucker, 2014); (c) GIMMS LAI3g product , which was further produced by GIMMS NDVI3g product using a neural network algorithm (Zhu et al, 2013); (d) PKU GIMMS NDVI (1982NDVI ( -2020, which is a new version of GIMMS NDVI product produced by a machine learning model incorporating Landsat images (Li et al, 2023); (e) GLASS LAI product (1981, which has 8-day temporal resolution and 0.05°spatial resolution, and was reconstructed by combing AVHRR LAI from 1981 to 1999 and MODIS LAI from 2000 to 2018 using a bidirectional long short-term memory (Bi-LSTM) model (Ma & Liang, 2022); (f) GLOBMAP LAI (1982LAI ( -2019 dataset at a spatial resolution of ∼0.07°, covering the period from 1982 to 2019, has half-month (1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000) and 8-day (2001-2019) temporal resolutions, and was produced by establishing a pixel-level AVHRR Simple Ratio (SR)-MODIS LAI relationship (Liu et al, 2012); (g) MOD13C1 NDVI (2000-2020) product (C61), which has a 0.05°spatial resolution and a 16-day temporal resolution (Didan & Munoz, 2019); (h) OCO2 SIF product (2000 at resolutions of 0.05°and 4 days, which was generated from MODIS surface reflectance and OCO2 SIF data using a neural network approach…”

Exploring the Potential of Long Short‐Term Memory Networks for Predicting Net CO2 Exchange Across Various Ecosystems With Multi‐Source Data

“…NDVI, which is a normalized ratio of the near‐infrared (NIR) and red bands, is valuable data for detecting vegetation status (Yin et al., 2022). We use the PKU Global Inventory Monitoring and Modeling Studies (GIMMS) NDVI product (Li et al., 2023). During the process of plant photosynthesis, leaves absorb photosynthetically active radiation (PAR) and release the unused portion of the absorbed energy in the form of fluorescence, which is referred to as SIF (Verrelst et al., 2016).…”

Earlier peak photosynthesis timing escalates global wildfires

“…We also used the PKU GIMMS NDVI product from combining AVHRR and MODIS observations 66 , which provides quality control (QC) layer to check the quality of NDVI value. Thus, we excluded the pixels contaminated by cloud, snow and ice.…”