Liangzhi You

Potential decarbonization for balancing local and non-local perishable food supply in megacities

Ensuring urban food security while reducing carbon emissions from food systems is a key challenge. Food localization can reduce transport emissions; however, its role in agricultural production emissions reductions is unclear. Here, we explored the effects of localization of seven perishables, incorporating emissions from production and cold-chain logistics, in Beijing and Shanghai, China; determined decarbonization under different scenarios by increasing or decreasing the localization, with or without constrains, of each food category (balancing strategy). The results show that every 1% increase in the localization of vegetables, poultry, and aquatic products decreased 2020 emissions by 0.4–1.9 tCO2e, but for beef and lamb, it increased emissions by 0.2–2.9 tCO2e. Localization decreased cold-chain emission shares for all foods. The balancing strategy with constraints reduced emissions by 0.76 MtCO2e (5%) and 0.44 MtCO2e (2%) in 2020 in Beijing and Shanghai, respectively. Utilizing urban agriculture at all costs (i.e., without constraints) further reduced emissions by a factor of 3–4. Over 90% of Beijing’s emissions added by 2035 under the business-as-usual scenario were projected to be offset by the strategy. In Shanghai, the strategy could reduce emissions by an additional 0.44 MtCO2e. The results indicate that expanding imports of carbon-intensive ruminant meat to replace local production and reallocating urban resources to vegetables, poultry, and aquatic products could lead to more sustainable food supplies in megacities. Further development of cold-chain logistics is expected to reduce emissions in synergy with the balancing strategy. Our results could help inform better food system planning in megacities.

Evaluating the gendered credit constraints and uptake of an insurance-linked credit product among smallholder farmers in Kenya

A band selection method for consumer-grade camera modification for UAV-based rapeseed growth monitoring

Near-infrared (NIR) modification of low-cost cameras is considered an important method to acquire high-resolution NIR images on an unmanned aerial vehicle (UAV) platform. However, few studies have examined filter selection methods to modify consumer-grade cameras for UAV-based agricultural crop monitoring. This study addresses a key challenge: how to balance imaging quality with spectral sensitivity when selecting filters for the modification of consumer-grade cameras. To this end, the normalized difference spectral index (NDSI) and the ratio spectral index (RSI) formulations were used to calculate the spectral indices (SIs) from all possible combinations of any two center wavelengths in UAV hyperspectral data. The contour maps of the coefficient of determination (R2) between the SIs and ground-measured rapeseed LAI were then computed to automatically generate the broadband combinations with optimized center wavelengths and effective bandwidths for selecting filters on camera modification. Results showed that a consumer-grade camera (Nikon D7000) modified by the selected filters had performance comparable with a multispectral camera (RedEdge Micasense 3), but slightly worse than a research-grade hyperspectral camera (Nano-Hyperspec®) in terms of SIs for LAI estimation. In addition, the high-resolution images from the modified camera were processed to obtain accurate crop plant height information. The SIs coupled with plant height from the modified camera (rRMSE = 18.1 % for field 1 and 14.3 % for field 2) was found to perform similar to, and in some cases even better than, those from the research-grade multispectral (rRMSE = 17.9 % and 16.7 % for the respective fields) and hyperspectral (rRMSE = 18.8 % for field 1) cameras for UAV-based LAI estimation. The findings from this study indicate that the proposed camera modification method is feasible and adaptable to agricultural crop monitoring. Thus, appropriately modified consumer-grade cameras can be a cost-effective replacement for research-grade sensors to rapidly and accurately assess crop growth status.