Search

• A new method for photovoltaic suitability is proposed, to avoid subjectivity. • Different site preferences between ground-mounted and floating PV are found. • The weighting results of different subjective and objective methods are compared. • The future scenarios of photovoltaic land optimization are designed. Land suitability evaluation is vital for large-scale photovoltaic (PV) plant construction. However, subjectivity in previous methods affects result reliability. We propose a new method using structural equation models to develop PV suitability indices, and demonstrate its validity in Zhejiang Province, China. By combining theoretic design with data mining technique, our method measures the structured paths between the overall index and evaluation indicators, exhibiting good model fit, flexible and objective weighting. Our results suggest that the spatial distribution of targeted PV plants based on suitability indices largely differs from that of observed ones. Most suitable areas for PV construction in Zhejiang Province are situated along the eastern coast, northern plain and river networks. The estimated PV technical potential for areas with and above moderate suitability is approximately 1,600 TWh/year, equivalent to 830 million tons/year of carbon emission reduction. Our study provides PV development strategies for similar areas in China and beyond. [Display omitted] [ABSTRACT FROM AUTHOR]

The utilization of renewable energy is an essential way to address climate change and achieve "carbon neutrality". As one of the most promising strategies for energy development, photovoltaic (PV) power generation has gained significant attention. In order to scientifically estimate its ecological benefits and potential for carbon reduction, it is imperative to have an accurate and efficient understanding of the current spatial deployment status of such facilities. As a novel economic mode, agrivoltaic systems innovatively integrate solar power generation to agricultural activities. The corresponding systems can achieve the diversity of land use, resulting in the mitigation of the issue of poor land utilization (caused by traditional PV panels' large footprint), which demonstrates the comprehensive benefits of clean energy supply, food production, and adaptation to climate change. However, there is still a lack of a systematic methodology for accurately mapping large-scale centralized PV plants in agrivoltaic systems (CAPVs), while simultaneously assessing their carbon reduction benefits over their entire lifespan. In this case, a systematic methodological framework in Zhejiang Province, China has been established and applied in this study. Firstly, a sample database of CAPVs by visual interpretation and data enhancement against satellite images has been built. Secondly, the CAPVs has been mapped using Google Earth images and U-Net network. Finally, CAPVs' carbon emission reduction benefits have been estimated based on Life Cycle Assessment (LCA) and corresponding spatial distribution patterns have been analyzed. The findings from the remote-sensing mapping indicate that the total area of the obtained CAPVs has reached 57.40 km2, with the majority of these facilities situated in the eastern and northwestern regions of Zhejiang Province. The overall accuracy of the deep learning model has reaches 90.42%, with the Mean Intersection over Union reaching 76.96%. Based on results from the carbon emission and reduction assessment, the CAPVs in Zhejiang Province have the potential to generate a total annual carbon dioxide (CO 2) output of 0.94 Mt./year over their lifespan with the accompanying carbon reduction benefit of 2.58 Mt./year, while, the CO 2 emission payback period would be only 3.95 years. The implementation of this study could offer valuable methodological guidance and data references for researchers and policymakers involved in technical research and decision-making related to agrivoltaic systems. Moreover, this study contributes to the synergy of multiple Sustainable Development Goals (SDGs) that pertain to the renewal of energy structures, reduction of carbon emissions, and enhancement of food security. • A deep learning-based high-precision framework for remote-sensing mapping of CAPVs is proposed. • A LCA-based method for estimating the carbon emission reduction benefits of CAPVs is proposed. • Agrivoltaic systems show considerable ecological benefits compared to traditional thermal power. • Agrivoltaic systems help to mitigate or avoid possible conflicts between SDGs. [ABSTRACT FROM AUTHOR]