Future Renewable Energy Systems Under Changing Climate

该研究项目的总体目标是评估由于二氧化碳浓度增加而导致的气候变化对未来可再生能源系统的影响。环境条件的变化将通过分析地球系统模型模拟来量化，该模型模拟由通过一系列二氧化碳排放情景得出的辐射强迫条件驱动。最新的耦合模式相互比较项目（CMIP）中的模式将根据其再现当前和过去气候、遥相关、季节内到数十年时间尺度变化的能力进行选择。利用这些模型，将在全球范围内对太阳能、风能和氢能系统相关参数的变化进行评估，并利用这些结果确定区域热点。例如，适合未来可再生能源安装的地点。然后，将使用新技术将全球气候变化信息缩小到一些热点，以得出相关的当地气象参数。然后，缩小尺度的气象参数将被用作天气-电力模型的输入，以评估未来的可再生能源发电。从计算科学的角度来看，问题的一部分可以被视为一个机会，开发统计模拟器，以填充一个大型的气候变化模拟合奏，评估模型预测的不确定性。问题的另一部分需要应用机器学习算法来定位全球气候变化信息。最后，需要数据科学技术将天气信息转换为更温暖大气中的发电。与开发改进的预测模型平行的一个方面是应用这些技术来了解外部强迫条件的变化（例如，不同水平的二氧化碳增加）如何影响大型跨国公司的能源基础设施和收入潜力，例如壳牌公司，其主要资产分布在地球仪各地。这反过来将有助于更好地做出明智的业务决策，并支持壳牌在未来几十年成为净零排放能源企业的目标。该项目为学生提供了一个很好的机会，可以在先进的数据科学，机器学习和优化方法方面进行培训，并接触可再生能源技术和气候变化影响。

The overall aim of this research project will be to assess the impacts of climate change due increasing CO2 concentration on future renewable energy systems. Changes in environment conditions would be quantified by analysing Earth System Model simulations driven by radiative forcing conditions derived through a range of CO2 emission scenarios. Models from the latest Coupled Model Inter-comparison Project (CMIP) will be selected based on their ability to reproduce current and past climates, teleconnections, variability on intra-seasonal to multidecadal timescales. Using these models, an assessment of changes in parameters relevant to Solar, Wind and Hydrogen energy systems would be made on global scales.Using these results regional hotspots will be identified. That is, for example, locations suitable for future renewable energy installations. The global climate change information will then be downscaled using novel techniques to some of the hotspots to derive relevant local meteorological parameters. The downscaled meteorological parameters would then be used as an input to weather-to-power models to assess future renewable energy generation. From computational science point of view a part of the problem can be viewed as an opportunity to develop statistical emulators to populate a large ensemble of climate change simulations to asses uncertainty in model predictions. Another part of the problem would require applications of machine learning algorithms to localize global climate change information. Finally, Data Science techniques would be required to convert weather information to power generation in a warmer atmosphere. A parallel strand to the development of improved prediction models will be to apply these techniques to understand how changes to external forcing conditions (e.g. various levels of CO2 increase) impact the energy infrastructure and revenue potential for large multi-national corporations, such as Shell, with a wide variety of major assets spread around the globe. This in turn will lead to better informed business decision making and support Shell's goal to become a net-zero emissions energy business over the next few decades. This project represents a good opportunity for a student to be trained in advanced data science, machine learning and optimisation methods, as well as being exposed to renewable energy technologies and climate change impacts.