Improved modeling of gross primary production and transpiration of sugarcane plantations with time-series landsat and sentinel-2 images.

Authors: CELIS, JORGE - University Of Oklahoma; XIAO, XIANGMING - University Of Oklahoma; White, Paul; CABRAL, OSVALDO - Embrapa

Submitted to: Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/1/2023
Publication Date: 12/21/2023
Citation: Celis, J., Xiao, X., White Jr, P.M., Cabral, O. 2023. Improved modeling of gross primary production and transpiration of sugarcane plantations with time-series landsat and sentinel-2 images.. Remote Sensing. https://doi.org/10.3390/rs16010046.
DOI: https://doi.org/10.3390/rs16010046

Interpretive Summary: Gross primary production (GPP) is a measure of the amount of carbon dioxide assimilated by plants through photosynthesis. Estimating GPP accurately is important to understanding the carbon cycle, vital to agriculture and predicting climate change. Sugarcane croplands account for about 70% of sugar production worldwide and are essential in biofuel production. Monitoring and predicting GPP helps improve crop management practices and crop yield. Our research evaluates whether satellite images that are moderate or high resolution can be used to estimate GPP of sugarcane crops. We studied crop growth and GPP at two locations over time: Louisiana, U.S.A., and Sao Paulo, Brazil. We calculated daily GPP using the Vegetation Photosynthesis Model (VPM) and compared it to the GPP estimated using satellite images. Our results showed that seasonal vegetation indices from both moderate and high-resolution satellite images produced similar GPP results when compared to VPM as well as on-the-ground sensors. The results from these analyses lay a foundation for us to run the VPM model with satellite images and predict GPP on crops like sugarcane over the contiguous United States and the world.

Technical Abstract: Accurate estimation of gross primary production (GPP) of terrestrial vegetation is crucial to understanding the carbon cycle, vital in agriculture, and predicting climate change. Sugarcane croplands (Saccharum spp hybrids) account for ~70% of sugar production worldwide and are essential in biofuel (ethanol) production. Monitoring and predicting GPP help improve crop management practices and crop yield. Although multiple GPP products are available based on different methods, most of them are only at moderate spatial resolutions and do not account for the differences between C3 and C4 photosynthesis. Our research evaluates and compares the potential of satellite images from moderate spatial resolution (MSR) and high spatial resolution (HSR) in estimating daily GPP of sugarcane crops. We studied temporal dynamics of crop canopy and GPP at two sites with industrial production and equipped with eddy flux towers: Louisiana and Sao Paulo, Brazil. We calculated daily GPP using the Vegetation Photosynthesis Model (VPM) with separate treatment for C3/C4 photosynthesis pathways, vegetation index data derived from these satellite images, and in-situ weather data. Our results showed that seasonal dynamics of vegetation indices from both MSR images (MODIS) sensor (500-m) and HSR images (Landsat -30m-, Sentinel-2 -10m-) tracked well with GPP from the EC flux towers (GPPEC). The HSR enhanced vegetation index (EVI) had stronger relationship with GPPEC than the MSR EVI. Furthermore, GPP estimates from the HSR images correlated more strongly with GPPEC than the GPP estimates from MSR images, highlighting the importance of HSR GPP estimates over small fields and sugarcane plantations with multiple crops. The results from these site-level analyses lay a foundation for us to run the VPM model with HSR images and generate daily GPP estimates at the field scale (10-m or 30-m spatial resolutions) on crops like sugarcane over the contiguous United States and the world.