Real-time forecasting of algal blooms in reservoirs

藻类大量繁殖是全世界水资源管理的一个重大问题，而且管理成本很高（例如，按2003年的价格计算，英国每年的成本估计为5000万英镑）。供水公司面临着过滤器堵塞、味道和气味差等问题，在更极端的情况下，还面临着藻类毒素含量高的问题。有许多管理战略可以实施，通常是以被动的方式实施。因此，能够预测何时可能发生藻华是有利的。在先前NERC资助的项目（UKLEON-NE/I 007407/1; http：//www.ecn.ac.uk/what-we-do/science/projects/ukleon）中，已经以可接受的准确度对湖泊中的藻类水华进行了预测。这些预测是使用计算机模型进行的，该模型描述了天气预报下藻类群落的生长情况。为了实现这一目标，需要足够的数据才能运行模型，并能够通知我们模型何时提供了湖泊系统的良好表示。充分的数据可用性是预报系统的一个关键部分，而且可能很昂贵，因此需要平衡数据收集和建模的成本与藻类大量繁殖管理的成本。拟议项目的总体目标是确定水行业对水库藻类预报系统的要求，并确定这种系统可能的成本效益。将对照与数据收集、模型校准和实施有关的成本，评估与不同管理战略有关的成本信息。这些费用将根据以下情况而有所不同：-所需预测期（例如提前3天、5天或10天）的预测准确性。- 水库及其集水区的特点。- 可用于建立预测系统的历史数据水平。如果这样的系统被证明是具有成本效益的，则在英国、欧盟和世界范围内对供水管理产生积极影响的潜力在水质和成本节约方面都是显著的。

Algal blooms are a significant problem for water management worldwide and are costly to manage (e.g. costing an estimated £50 million per year in the UK at 2003 rates). Water companies are faced with problems such as blocked filters, poor taste and odour and, in more extreme cases, high levels of algal-derived toxins. There are a number of management strategies that can be implemented, often in a reactive way. It is therefore advantageous to be able to predict when an algal bloom is likely to occur. In a previous NERC funded project (UKLEON - NE/I007407/1; http://www.ecn.ac.uk/what-we-do/science/projects/ukleon), forecasts of algal blooms in lakes have been made with acceptable accuracy. The forecasts are made using a computer model which describes the growth of algal communities given weather forecasts. For this to be achievable, adequate data is required to be able to run the model and to be able to inform us of when the model is providing a good representation of the lake system. Adequate data availability is a critical part of the forecasting system and can be costly, so there is a requirement to balance the costs of data collection and modelling against the costs of managing algal blooms. The proposed project has the overall objective of defining the water industry's requirement for an algal forecasting system for reservoirs and to determine the likely cost-effectiveness of such a system. Information on the costs associated with different management strategies will be assessed against the costs associated with data collection, model calibration and implementation. These costs will vary based upon: - The accuracy of the forecasts for the required forecast period (e.g. 3, 5 or 10 days ahead). - The characteristics of the reservoir and its catchment. - The level of historic data available for setting up the forecasting system.If such a system is proven to be cost effective the potential for positive impacts on water supply management within the UK, the EU and world wide are significant both in terms of water quality and cost savings.