Nick J. Mount's research while affiliated with University of Nottingham and other places

... ISM is a dependable method for hierarchical structural modeling, surpassing other models (Maskrey et al., 2021;Kyriakarakos et al., 2012). Combining MICMAC and ISM generates privileges for decision-makers (Manjunatheshwara and Vinodh, 2018). ...

... Different weighting techniques have been tested for combining hydrological models' outputs at basin and global scales (Arsenault et al., 2015b;Wan et al., 2021;Zaherpour et al., 2019), showing that weighted multi-model ensembles generally present more robust performances than single hydrological models. Most of these studies have been carried out over historical streamflow records and have particularly focused on forecasting applications. ...

... Collaborative modeling can be in the form of undertaking workshops and/or focus group discussions along with decision tools to address the issue. Such examples of decision tools applied in different fields, as investigated by Maskrey et al. (2021) [25], are the use of Fuzzy Cognitive mapping in water resources management [36], System Dynamics to reduce levels of vulnerability and exposure [37], and Bayesian Network to investigate the sensitivity of measures [34]. The mentioned techniques have their strengths and limitations, and this must be recognized by considering its characteristic of being holistic, adaptable, accessible, evaluative, and transparent. ...

... Hydrologic rainfall-runoff models are often used in PUB, including conceptual models (Pelletier & Andréassian, 2022), physical-based models (Gou et al., 2021), and data-driven models (Yu & Ma, 2021). Regional physical-based models describe hydrological processes within catchments more accurately than conceptual models but are less efficient and less scalable (Bierkens, 2015;Zaherpour et al., 2018). Data-driven models encourage hydrological modeling by identifying input-output relationships with greater precision and efficiency but fail to describe physical processes (Bergen Karianne et al., 2019;Lin et al., 2021;Shen, 2018). ...

... 1) It is imperative that data sets used for model development and evaluation are different. To enable the true generalization ability of models (knowledge of students) to be tested, rather than how well they have memorized the input-output relationships in the model development data (homework questions), it is important that different data (questions) are used for model development (homework) and evaluation (exam) Humphrey et al., 2017). While teachers can use their experience and understanding to generate a wide variety of representative homework and exam questions, modelers typically only have access to a fixed data set, which therefore needs to be split into development and evaluation subsets. ...

... Scientific efforts in improving LSMs are often, however, confined to a single biome or process. For example, LSM benchmarking studies may focus on temperate broadleaf phenology [22], dryland phenology [23], large river hydrology [24], groundwater storage [25], and more. Yet, studies of ecosystems and landscapes often take a more synoptic approach, considering all or many of the flows of materials and energy that enter and leave a geographic area [26]. ...

... Studies have demonstrated hydrological responses that vary spatially across the globe (e.g. Arnell and Gosling 2013;Gosling et al. 2017;Do et al. 2020), for different parts of the same catchment (e.g. Thompson et al. 2014Thompson et al. , 2017a, and contrasting impacts on high, mean and low flows (Giuntoli et al. 2015;Chan et al. 2020). ...

... On the other hand, numerous collaborative modeling approaches attempted to develop a decision support tool and set up a model by including the local knowledge and perceptions of the stakeholders [32,33]. Moreover, the engagements of stakeholders in the development of FRM and water resources management intervention options are also explored in some studies [12,34,35]. ...

... Artificial intelligent (AI) algorithm-based models are flexible for tuning the model structure and can learn the historical patterns from the data by optimizing its parameters (Mount et al., 2016). AI-based models are known as a black box model that is a system using inputs and outputs to enhance the predictability without any knowledge of its internal workings (Castelvecchi, 2016). ...