【作者】 石洪华；

【导师】 丁德文；

【作者基本信息】 中国海洋大学 ， 环境工程， 2008， 博士

【摘要】 生态风险是指一种或多种应力(物理、化学以及生物应力等)作用而导致生态系统某些负面生态效应发生的可能性。生态风险评价是对生态风险引起的负面生态效应发生概率进行评价，并实施风险管理、防止或降低风险的过程。其目的是通过风险评价，为风险对策提供科学的信息，以便使污染或其它的一些生态环境损害降到最低程度。早期的生态风险评价，绝大多数是针对于人类(或人类环境)所受的风险进行的。直到最近几年，由于全球生态环境的恶化，针对生态系统的风险评价才得到逐渐的重视。作为人类活动最集中的场所，城市是一类典型的社会-经济-自然复合生态系统，应成为生态风险研究的主要对象。城市生态问题的产生包括自然因素，如全球变化、地壳运动等；更主要的是人类自身生产、生活过程中造成的干扰。城市化是区域可持续发展的重要步骤，然而城市化过程与生态环境之间存在着各种各样的矛盾与胁迫，同时也伴随着生态风险的产生，评价城市化过程中诱发的生态风险，并找出各种生态风险变化的规律性，对城市未来的健康发展具有重要的现实意义。城市污染、气候变化(如洪涝灾害等)等生态风险是当前城市发展的重要瓶颈。当前关于区域生态风险评价的研究主要集中在自然生态系统，对于城市的生态风险评价研究并不多见。本文以典型滨海城市——青岛市崂山区作为研究区域，系统剖析该区域典型生态风险，研究这些风险的特征及变化趋势，根据可靠性数学理论构建生态风险评价指标及其计量模型，对典型滨海城市空气污染、气温改变、降水量变化等生态风险进行评价，深刻剖析了其生态风险特征，进行评价，并根据均生函数、人工神经网络等现代数学工具开展预测研究。在国际流行风险管理模型VaR的基础上，构建了区域生态风险管理的EVR模型，并开展了应用研究。考虑到城市生态系统的主体——人类对生态系统变化(风险)的响应，作者基于滨海城市这一人类生态系统的特征，构建了城市生态适宜度的数学模型，评价了研究区的生态适宜度及其演变趋势。最后，作者开发了滨海城市生态风险评估的决策支持系统软件，作为滨海城市生态风险管理的重要手段。主要研究内容与结论如下：(1)根据可靠性数学思想，用风险度、可靠性指数、风险指数、系统持续稳定性指数、系统恢复性指数、整体损失度等指标构建了城市生态风险评价模型。该模型具有指标意义明确、计算便捷、针对性强等特点，并在崂山区开展了应用研究。(2)根据生态风险评价模型，对青岛市崂山区2001-2005年空气污染生态风险及其演变趋势进行了研究。研究发现，2001-2005年间，多项风险评价指标均未见差异。但值得注意的是该区空气污染的恢复性指标明显提高。从2001年的45％，2002年的40％上升至2004年的60％、2005年的50％。这很大程度上反映了该区污染治理的力度加大；应对突发污染事件的能力增强，治理污染速度加快。同时，该区中度以上污染事故的发生率也明显减少。这是该区生态质量改善的重要标志之一。因此，整体损失度也呈下降趋势，2004年降为6．44％。冬春两季仍是空气污染事故的高发期。5年来，冬季空气质量状况并未好转，反略有进一步滑坡的趋势。应当引起足够的重视。(3)用趋势分析、相关分析等方法对崂山区近50年降水资料进行分析，用极值估算、均生函数等方法建立了降水预测模型。对崂山区2005-2010年的年降水、各季节降水、旱涝等级、夏季暴雨频数等进行了预测。预测结果与2006年的实况基本吻合。研究发现：崂山区降水量季节变化较大，降水主要集中在夏、秋两季，冬、春两季降水明显偏少。年降水量的变化近20年来有趋稳定的势头，但夏秋两季有逐年下降的风险。(4)研究了崂山区50年来气温变化趋势，发现该区发现，该区的气温升高主要发生在近20年间。1954-1975年共20年时间，未见明显的气温升高证据。而在1 981-2004年均升高气温最低在0．035℃，最高达到0．092℃，24年间平均每年升高0．052℃。这明显高于东南沿海地区平均水平。作者还建立了多输出的2类人工神经网络模型预测了未来10年温度变化趋势。(5)青岛市崂山区生态适宜度调控对策研究计算结果表明，1996-2004年，青岛市崂山区人类生态适宜度平均为57．40％，高于全国平均水平(50％)。同时，我们注意到，该区的人类生态适宜度在经历了一个短期回落以后，已进入快速增长的时期。特别是近3年来，崂山区的生态适宜度均远高于全国平均水平，充分显示了该区强劲的综合竞争能力和可持续发展能力。(6)构建了生态系统风险管理的EVR模型，并基于Monte Carlo方法，以青岛市崂山区为例开展了空气污染状况案例研究。(7)开发了“青岛崂山生态风险决策支持系统”。作者认为在城市生态风险评价中，应要充分体现人的主体地位。建议按照驱动力(D)-压力(P)-状态(S)-响应(R)-调控(C)的系统工程框架，开展区域生态风险评价，该框架是对当前广为使用的压力(P)-状态(S)-响应(R)框架的扩充。以人口迁移为主要指标，根据人口迁移的动力学机制，构建了人口迁移势、生态适宜度模型，通过分析区域人口迁移的特征，巧妙地利用少量的数据研究了崂山区生态适宜度变化趋势。另外，本文构建的生态风险管理模型EVR充分借鉴金融市场风险管理的最新理念，对于区域风险管理具有重要参考价值。更多还原

【Abstract】 Ecological risk refers to the possibility of some adverse ecological effects due toone or more stress (physical,chemical and biological stress,etc.) Ecological riskassessment is the whole process,include conduct probability assessment of adverseecological effects,and implement risk management to prevent or reduce the risk.Itspurpose is to provide scientific information for risk management,and minimizedamage of pollution and ecological degradation.Early ecological risk assessmentfocus on the risk that human (or human environment) suffer.Recently risk assessmenton ecosystem attracts more and more attention due to the degradation of the globalenvironment.As most activities are centralized in the city,city is a typical complex ecosystemwith society system,economy system and nature system,which becomes the mainresearch target of ecological risk assessment.The reason causing urban ecologicalproblems include natural factors,such as global change,crustal movement etc.Anddisturbance from human production and daily life is another important reason.Theurbanization is an important step to realize regional sustainable development.Butthere are various conflicts and stress between the urbanization and ecologicalenvironment and ecological risk will occur simultaneously.It plays an important rolefor urban sustainable development to assess the ecological risk in the urbanizationprocess,and find out the laws of ecological risk.Urban pollution,climate change (such as floods etc.),and other ecological risk isthe major obstacle for urban development.At present the study about the regionalecological risk assessment mainly focus on natural ecosystems,the studys on urbanecological risk assessment are rare. As a typical coastal city,Laoshan district in Qingdao is selected as the study area.The ecological risk and the character in this area are analyzed.According to reliablemathematical theory,ecological risk assessment indicators and models are established.And the ecological risk is estimated including air pollution,change in temperature andchange in precipitation in Laoshan district.The characters of the above risk arepresented and forecast research is conducted according to Mean Generation Function,artificial neural networks and other modern mathematical tools.Based oninternational VaR model,EVR model of ecological risk management is establishedand it is applied in Laoshan District.Considering the response of human to ecologicalrisk and the character of coastal ecosystem,niche-fitness model is established and it’sused to assess the niche-fitness of the studied area and its trend.Finally,decisionsupport system software of ecological risk assessment in coastal city is developed andused for an important means for ecological risk management.The main results and conclusions are as follows:(1)According to reliable mathematical theory,urban ecological risk assessmentmodel is established including the following indicator:risk degree,reliability index,risk index,system continued stability index,system restoration index and the overallloss.This model has the character of clear indicators,calculated convenient,well-targeted.And application research is conducted in the Laoshan District.(2)According to ecological risk assessment model,air pollution and ecologicalrisk evolution in Laoshan district of Qingdao City from 2001 to 2005 is studied.Theresult indicates that a number of risk assessment indicators are no differences from2001 to 2005.But it is worth noting that the restoration index of air pollution in thisarea increases significantly.It increases from 45% in 2001,40% in 2002 to 60% in2004 and 50% in 2005.This reflects that more efforts in pollution control,theimprovement of capacity to deal with the unexpected pollution incidents,and thequick treatment for pollution.At the same time.the incidence of moderate pollutionaccidents also decrease significantly.which indicates that ecological quality improves.Therefore,the overall loss show a downward trend.reduced to 6.44% in 2004.In winter and spring,there is still high incidence of pollution accidence.Over the pastfive years,air quality in winter has not improved,but there is the trend of furtherlandslides.More attention should be paid to this problem.(3)The rainfall data in the past 50 years is analyzed using trend analysis,correlation analysis and other methods.And the precipitation forecast model isestablished using extreme estimates method,Mean Generation Function etc.Annualrainfall,the seasonal rainfall,drought and flood levels,frequency,such as the summerstorm in Laoshan district are forecast from 2005 to 2010.The forecast results arebasically consistent with the facts in 2006.The result shows that there are largerchanges in precipitation season in Laoshan District,and precipitation is mainlyconcentrated in the summer and autumn,and in winter and spring precipitation issignificantly less.Changes in precipitation in the past 20 years show a more stabletrend,but precipitation in summer and fall has decreased year by year.(4) The change in temperature in the past 50 years is analyzed.The result showthat the temperature increase occurred mainly in the past 20 years.From 1954 to1975,there is no obvious evidence of warming.But from 1981 to 2004 the minimum of theaverage annual increase in temperature is 0.035℃,the maximum is 0.092℃,andaverage annual increase is 0.052℃.This is obviously higher than the average level ofthe southeast coastal areas.A multi-output of two kinds of artificial neural networkmodel is established to predict the temperature trends in the next 10 years.(5) The result of niche-fitness research in Laoshan district shows that the averageniche-fitness is 57.40%,higher than the national average (50%).At the same time,wenote that the niche-fitness in this area has entered a period of rapid growth after ashort-term decrease.Especially in the last three years,niche-fitness are far higher thanthe national average,fully shows the area has strong competitive ability and capacityfor sustainable development.(6) Establish EVR model of ecological risk management,and conductapplication study in Laoshan district based on the Monte Carlo method. (7) Develop“Decision support system software of ecological risk assessment inLaoshan district of Qingdao”It is necessary to fully reflect the dominant position of human in urban ecologicalrisk assessment.It is recommended to conduct regional ecological risk assessmentaccording to the system engineering framework:Driver (D)-Pressor (P)-State(S-Response(R)-Control(C).The framework is an expansion of the current widely usedpressure (P)-Status (S)-Response (R) framework.Take population migration as themain indicators,human ecological momentum and niche-fituess model is establishedaccording to the dynamics mechanism of population migration.Based on analysis ofthe characteristics of regional population migration and clever use of small amountsof data,the change in niche-fitness in Laoshan district is analyzed.In addition,EVRmodel of ecological risk management fully draw the concept of risk management onthe financial markets,which is very important for regional risk management.更多还原