AMAZING- Advancing MAiZe INformation for Ghana

点击翻译按钮获取中文摘要

The agricultural sector is under increasing pressure from population growth, climate change and environmental issues such as soil erosion, drought, flooding and pesticide overuse. In China, which has 20% of the world's population and only 10% of the arable land and water resources, huge changes have been made in agricultural practice in recent years to maintain food security. In the North China Plain (NCP), 61% of China's winter wheat and 45% of its maize are grown, but this is dependent on irrigation, causing a 1 m annual drop in the water table. Since 1985, double cropping of winter wheat and maize has been used to increase supply but this exacerbates water issues even further. China is investing heavily in technology at farm, regional and global scales and remote sensing forms a vital part of that strategy, for monitoring crop health and production.In Ghana, more than half of its labour force is engaged in agriculture, which contributes to 54% of its GDP and provides over 90% of the country's food needs. At the same time, Ghana's agriculture predominantly involves smallholders and relies heavily on rain-fed subsistence farming practices, resulting in food production being below its potential. Similar to many other countries in the region, smallholders are widely considered to be the most vulnerable component of Ghana's rural sector. The project will use earth observation (EO) data and crop modelling techniques to help Ghana to build a national crop monitoring capability, informed by state-of-the-art developments in EO and crop modelling. Among various types of crops, food-crop farming is vital to low-income households and particularly women and children relying on subsistence agriculture. Timely monitoring information will result in better informed farmers and extension workers, as well as government, which will lead to better evidence-based decision-making, more efficient farming management, and more sustainable development in the Ghanaian agricultural industry in the long run. Field survey and census has been traditionally used in many countries for crop yield estimation, but these have gradually been replaced or supplemented in many countries by EO-based estimates. EO data can provide a frequent measurement of a number of crop relevant biophysical parameters at a high spatial resolution (10-20 m). Using these data to monitor croplands, however, is hampered by the inherent difference between the nature of EO measurements and the agronomic parameters of interest. An alternative but equally compelling line of research is that of physical-based crop growth models (CGMs), which predict the crop development as a function of meteorological drivers, soil properties, specific cultivar parameters and management practices. The predictions from these CGMs, however, are not capable of describing a particular field with detail due to insufficient information on management practices, microclimate and soil variations, pests, etc. One way to account for this local variability is to use satellite observations to "correct" the model trajectory. In these "data assimilation" (DA) systems, the model "learns" about the reality of the crop from the observations. The UCL team and its collaborators have conducted a series of DA work in China and Ghana in recent years. From these pioneering works, a number of major research challenges have been conquered for implementing a DA-based crop monitoring system. This project will demonstrate the practical benefits of EO-enabled crop monitoring and yield prediction for sustainable agriculture, to continue and deepen our collaborative partnerships between the UK and China and extend its impact, as well as to develop a partnership with Ghana to adapt the approach to meet local needs and conditions. This project will also present an opportunity to train in-country specialists both in EO, physical modelling, data assimilation, as well as in standard geoprocessing techniques.