Marine aquaculture areas in a region up to 30 km from the coast in China were mapped for the first time. It was found to cover a total area of ~1100 km², of which more than 85 % is marine plant culture areas, with 87 % found in four coastal provinces. The results confirm the applicability and effectiveness of deep learning when applied to GF-1 data at the national scale, identifying the detailed spatial distributions and supporting the sustainable management of coastal resources in China.

REDD+ requires forests to be measured. However, many communities have their own concepts of forest with different meanings. Global forest inventories frequently ignore these conceptualizations. This paper describes an approach for generating alternative measures of forest simultaneously to support the international objectives of activities such as REDD+ and to reflect local concepts and semantics associated with forest.

This manuscript looks at the measurement of water quality parameters through the collection of physical samples which are then analysed in an analytical laboratory compared to data generated by equipment placed in the environment. The advantage with this is that large volumes of data are collected at relatively low cost, however the instruments are not subject to the same scientific rigour as in the laboratory. A comparison of the two allows an assessment to be made of the two approaches.