荷兰沿海及莱茵河三角洲区域洪水风险管理的近期发展

荷兰沿海及莱茵河三角洲区域洪水风险管理的近期发展 黄 波 山东黄河勘测研究院，济南 摘要: 在1953年由海暴潮引发的灾难性洪水和1995年的莱茵河洪水之后，荷兰在沿海及沿莱茵河区域均实施了大规模的堤防建设和堤防加固工程。洪水安全防御标准目前已经达到从东部的1/1250到沿海的1/4000和1/10000。然而随着将来天气变化的不确定性，以长远的观点，荷兰对于洪水管理策略的争论变得更加广泛和激烈。近期对莱茵河洪水流量的估测显示，设计流量预计将从20世 33纪90年代的15,000 m/s 增大到2100年的16,800(最小)或18,000m/s(最大)。另外，在下游三角洲区域，海平面上升可能对洪水下泄产生阻碍，预计到2100年海平面将上升0.2到1.1m，在三角洲及冲积平原由于泥炭土层的压缩和氧化将最终导致大面积的地面沉降。 传统的大堤加高及加固措施在防洪方面是有效的，但是随着将来的社会发展，相对于洪水安全，对空间功能的需求可能越来越多，例如景观、环境、生态、居住、工业等等。因为这些因素，传统的大堤加固方式将引起社会越来越多的反对，这就要求在洪水风险管理方面需要策略上的转变，即综合河流管理和海岸区域空间规划。 1. 背景介绍 1953年的海暴潮是荷兰历史上最大的自然灾害之一，死亡人数1853人，是荷兰历史上一个史无前例的数字，令人警醒、难忘。在此次海暴潮中发生在Hoek van Holland的洪水水位是目前三角洲北部的标准水位，高于平均海平面3.85m是曾经出现过的最高水位，比历史记载中1894年的最高水位3.28m还高出0.57m。 灾难发生之后，荷兰开始了大规模的堤防建设，其中的三角洲工程将沿海所有的出海口进行封闭，而且沿着海岸及河流下游区域所有防洪工程均进行了史无前例的加固，使得该区域的防洪水平第一次达到了设计标准。在中央,荷兰地区防洪工程达到了抵御10,000年一遇风暴的防御标准。 在1993和1995年，莱茵河水位暴涨，荷兰经历了一段惊魂未定的时期，河水流量在Lobith达 33到了12,000m/s，在记载中只有在1926年(12,600m/s)莱茵河水超过了此流量。在洪水期间，由于大堤稳定无法得到保障，超过240,000人被迫从多个迂田进行撤离，幸运的是大堤没有决口。 此后，从1995年开始实施工程性措施，一项被称作“三角洲大河计划”的紧急法案获得议会通过，该法案要求所有沿河大堤的加固项目必须在2000年前完成。基于此，所有沿河大堤很快按照 31250年一遇的防洪标准进行了加固。目前，沿河地区可以在河水流量在Lobith达到15,000m/s时得到有效保护。 1995年的莱茵河洪水之后，联系到将来天气变化的因素，在荷兰关于洪水控制策略的讨论愈加 3广泛和激烈。近期对莱茵河洪水流量的估测显示，设计流量预计将从20世纪90年代的15,000 m/s 3增大到2100年的16,800(最小)到18,000m/s(最大)。另外，在下游三角洲区域，海平面上升可能对洪水下泄产生阻碍，预计到2100年海平面将上升0.2到1.1m;在三角洲及冲积平原由于泥炭土层的压缩和氧化将最终导致大面积的地面沉降，在荷兰过去1000年的排水过程中地面沉降过程一直保持持续和加快。 总之，由于洪峰流量的增大和海平面的上升河流设计水位可能会升高，然而在被保护区域由于人口增加和经济发展其脆弱性也在增大，而且土地下沉会使形势更加恶化。这就要求在洪水风险管理方面需要策略上的转变，即综合性河流管理和海岸区域空间规划。本文首先对当前莱茵河沿岸及海岸的情况进行简单介绍，然后对当前在洪水风险管理方面存在的问题进行分析，最后提出相应的解决措施。 1 2. 洪水风险问题 2.1 莱茵河水流量增大 由于全球气候变暖，莱茵河流域的降水方式也将会发生改变。预计莱茵河将从目前的以雨水和冰雪融水为水源逐渐转变为以雨水为主的河流，冬季流量大而夏季流量小。由于冬季降雨量的增加将导致莱茵河冬季流量的增大，而夏季流量则由于冰雪融水量的减少和蒸发量的显著增加而减小，在夏季蒸发量的增加会超过平均降雨量的微弱增长对河流流量的影响。 四种预测天气变化(2100)的河流预期流量 (左:马斯河 右:莱茵河) 上显示在所有天气变化预测中(低、中、高降水、干旱)，相对于目前的流量，预期冬季河水流量将增加，而夏季流量将减小。 2050年河流流量预测值(左:莱茵河 右:马斯河) 在《防洪法》中设计流量被作为法定安全标准的基础，代表所能防御的最大流量。在莱茵河/马斯河地区，设计流量是基于一个流量值，其平均每1250年出现一次。在法律条款中，它是在不发生洪水的情况下河流所必须能排泄的最大流量值，大堤、滩区、主河槽以及相关因素均由该流量确 3定。随着在1993和1995年莱茵河/马斯河出现的高水位，莱茵河的设计流量从15,000m/s调整到 2 316,000m/s。这次调整的直接原因就是当前的形势无法满足法定的安全标准。 根据气候变化的预测方案，在2050年莱茵河和马斯河的设计流量都将增大:莱茵河约增长3-10%，马斯河约增长5-20%。这就意味着如果要确保达到法定的安全标准，采取一些其它辅助措施是很有必要的。 2.2 海平面上升 预期到2100年，在荷兰沿海因气候变化将导致海平面相对于目前地面高程上升20-110cm。该预测是基于地面沉降值为每百年10cm，并且同时考虑上个冰川纪对地面沉降的后续影响(NAP荷兰标准海平面高程也受其影响)以及由于泥炭和粘土层沉积造成的地面下沉平均值。然而，在不同地区可能发生明显不同的地面沉降。要知道在海水温度变暖和气温升高之间有一个相当大的时间滞后期，这就意味着如果由于废气排放的减少导致大气平均温度的升高受到限制，则海平面的上升只能是几个世纪以后的结果。 预期海平面上升的后果之一就是需要向海岸滩地增加大量的沙以补充目前正在发生的沙的流失，以保持当前的安全标准。对海岸系统来说，沙的补充可以确保海岸、河口及Wadden海与海平面上升保持同步。在将来，需要更宽、更坚固的大堤来抵消因海平面上升造成的海水压力的增加。 在将来的50年，海岸管理的附加成本预计不超过GDP的0.13%，但随着2050年之后海平面的进一步上升，海岸管理的成本可能会远远超过当前的费用水平。除了海平面上升的影响，洪水水位还很大程度上取决于发生于北海的风暴，然而，风暴的频率和强度在将来如何变化依然未知。 2.3 地面沉降加快 除了气候变化和地壳运动引起的地面沉降，荷兰还面临泥炭土层地区的沉降问题，从中世纪起在泥炭土层区域的沉降已达2-3m。该沉降与泥炭的随水外排密切相关，同时其自身萎缩以及氧化并以CO进入大气，沉降值最大可达每年1cm，这依赖于水位的变化。如果按照这样的速度，到2 2050年沉降值将达到0.5m。如果地面沉降持续下去，对于含有深厚泥炭土层的地区，尤其是荷兰西部含有厚达12m的泥炭土层，以长期来看，洪水影响、地表水盐化问题会加剧，水管理的难度也会加大。在其他几个地区(如在荷兰东北部Slochteren附近)油气开采也会引起地面沉降，预计在此地区会产生额外60cm的沉降。温度的上升、夏季变长、干湿状况的巨大差别很可能会导致泥炭土氧化的加快，反过来会加速地面的沉降。 由于泥炭土壤的变化和水管理的差异地面沉降的速率在各地是一样的。例如，农业区需要相对深的排水水位而在属于泥炭土层的市区则需要相对高的地下水位以防止木桩基础的腐化。因此，在属于泥炭土层的地区，水管里系统变得愈加难以统一，土壤盐份的渗出加剧(对农业有害)，而且道路及建筑物的下沉也造成诸多破坏。因此，各个省份(尤其在荷兰西部)采取了一些相应的措施，然而，这些地区只占泥炭土总面积的4,。 2.4 海暴潮危害 从1962年起海洋风暴的数量在逐渐减小，下图显示了荷兰在过去41年700次大型风暴的分布情况。这些风暴的风速取决于在荷兰所处的位置，一般超过11-16m/s，按蒲福氏风级这相当于6-7级的风力。另外，即使只考虑300或500次极大型风暴，分布图也没有变化:风暴的数量在减少。至于数量的减少与温度上升之间在多大程度上存在关联依然无法确定。 在荷兰天气变化对风暴形式影响的不确定性意味着对风暴型洪水的可能性变化还没有足够的认识。当前，通过大型的模型研究显示“超级风暴”发生的概率依然存在，其可能的风速会远远超过荷兰在20世纪所经历的风暴。因此，为了认识其潜在的过程，有必要通过更精确的模型进行进一步研究。 3 荷兰大型风暴分布图 3. 洪水风险管理措施 3.1 还河流更大空间 沿莱茵河分支河流采取滞洪措施 滞洪就是削减一段洪峰并将削减的水量暂时储存于一片被大堤围成的区域，一旦险情过去就将蓄存的水量再次排除，这样，滞洪区就限制了排向莱茵河下游分支(一个或多个)河流的洪峰流量。这也就意味着如果要充分发挥滞洪区的功能，滞洪区就应该尽可能位于上游区域，对荷兰来说就是应该尽可能在Lobith(莱茵河流入荷兰的第一个小镇)附近。 33滞洪区的概念也可以这样理解:目前的设计流量15,000 m/s与预期设计流量16,000 m/s之间的 33差值为1000 m/s，如果通过滞洪的方法来避免进一步加高大堤，那么这1000m/s的水量就必须引入滞洪区，所需要的总蓄水量由洪峰水位与设计水位之间的差值来确定(特别是洪峰的持续时间)。通过这种方法就可以计算究竟有多少水量需要暂时蓄存于一个还是多个被大堤围成的区域。 3对于上面所提及的1,000m/s的流量，如果洪峰以均值持续几天，那么需要蓄存1.7至2亿立方米的水量。如果以平均5 m的水深，则需要3,500至4,000公顷的土地面积，这样的水深在Boven-Rijn 和Waal 河段是可以达到的，但是沿着Neder-Rijn 和Ijseel 河段实际水深要相对更浅。滞洪区域所处位置愈高，则蓄水深度就愈浅，所需的蓄水面积也就愈大。 增大莱茵河分支河流泄洪流量 相比于滞洪措施，两者最大的不同在于滞洪是减少泄水量，而增大河流泄洪容量则是在保持同样流量的情况下降低洪水水位。如果蓄水能够保持尽可能长的时间，对滞洪区下游地区或在上游较 4 短的区域是有益的，而增大泄洪容量则对实施河段的上游地区有益，原因就是由河道缩窄、河道障碍或较大河道糙率造成的壅水效应被大大减小或河道断面被增大。增大河流泄水容量的措施必须首先在下游进行实施，随后在上游方向采取相应的措施。 在增大泄水流量方面有很多措施，我们可以将它们主要分成三类: , 针对主河槽(小流量河槽)采取的措施; , 针对滩区采取的措施; , 在大堤以外采取的措施(如大堤后撤) 在每类措施当中都有如下一些具体的方法和措施。 , 降低主槽河床 通过计算显示采用疏浚降低河床可以在50km范围内降低20到30cm的水位。 , 降低丁坝 通过降低丁坝高度可以使更大水流通过丁坝而使主槽流量减小，这样能减少或推迟水流对主槽的侵蚀。丁坝降低对减少一些对主槽不利的侵蚀是一项比较可取的措施。 , 滩区开挖 滩区开挖可以应对因沉积而造成的滩区的逐年抬高，该措施可以与近年大堤加固中的粘土采挖相结合或者与自然环境的改善相结合。 , 消除水力瓶颈 5 首先要基于三条莱茵河分支河流的水位坡度线来确定何处为水力瓶颈，通过这种方法共发现了254处。然后，借助地形图确定它们的具体地点和类型，如禁淹区、桥墩、渡口坡道等。某些水力瓶颈包含大面积的地表区域，如禁淹区(工厂)，而像桥墩、渡口坡道或夏季坝(主槽小坝)则要小的多。 不同的措施对降低水位的作用和效果差异明显，基于这点，确定了两项标准来选择60项具体措施(包括18个小规模的大堤移位)，这些标准仍需进一步的研究，即: , 水位降低效果至少1cm(否则就不值得实施); , 水位降低效率至少2mm/百万欧元(否则过于昂贵); , 大规模大堤后移 大堤后移措施相当昂贵，但是却非常有效。通常工程成本从400万到5000万不等，但是大堤后移的效果均超过每百万欧元降低2mm水位的水平。单纯大堤后移所获得的降低水位效果可以达到几十厘米，而此类其他措施只能降低几厘米，同时该措施也满足消除水力瓶颈所确定的标准。 在滩区缩窄而引起上游长距离壅水的情况下，大堤后移的效果尤其明显。这就是为什么大堤后移会对上游长距离降低水位产生作用的原因。每项措施通常都能实现降低当地水位10-20cm。在沿Waal河和Neder-Rijn/Lek河段，所有大堤后移措施可以最大降低水位60cm。 , 针对城区瓶颈的“绿色河流”措施 对上游窄河段的加宽或挖深并不能完全解决问题，因为这就像在“水库”内或沿着“水库”进行加宽或挖深而无法打开出水口排出“水库”内的水一样。在下游采取类似的措施对降低水位的效果非常小，这就好比用一个堵塞的阀门进行抽水。城区造成的瓶颈不但使此类措施毫无帮助，对其它措施也产生危害，而且根据综合措施成本还可能会产生更不利的作用。 6 显然，要消除由城区瓶颈带来的影响只有在坝外区域采取措施，毕竟在此区域几乎没有滩区。以上提到的一些措施在早期阶段并没有加以考虑，因为在建筑稠密的地区实施这些措施是不可行的。鉴于这个原因，在涉及城区瓶颈的区域实施了“绿色河流”的措施。 绿色河流实际上是在两个导流堤之间的滩区，当河流流量小时没有水流通过，只有在发生洪水时才发挥作用，它可以用作农业或者设计为自然、休闲娱乐区，总之为“绿色”。 3.2 沿海区域综合性防洪策略,ComCoast 概念及目标 随着预期气候的变化，全球沿海地区的防洪设施将承受越来越大的压力。荷兰一直以来通过传统的加高大堤的方法来防御日益增长的洪水威胁。然而，随着海平面的持续上升，越来越多的证据表明抵御洪水不能只简单依靠加高大堤而应该寻找其它更具创造性的策略。ComCoast就是这样一种创新性的洪水风险管理策略，它通过在海、陆之间的逐渐过渡来建立一道综合的洪水防御区域，它同时具有广泛的环境功能，如休闲、渔业、旅游业和自然景观。 沿海区域综合性防洪策略的目标就是通过在海陆之间的逐渐过渡来创造多功能的洪水管理方法，在提供更多经济发展机会的同时，使广大的沿海社区及环境受益。ComCoast的目标为: , 在沿北海区域，为沿岸防洪策略探索在目前或将来能发挥空间潜能的地点或区域; , 从经济和社会角度创造和实施新的方法对多功能防洪区域进行评价; , 结合环境、人口以及在确保所需安全水平的情况下创造和发展新的防洪技术措施; , 以公众参与为重点改善和实施利益共享者合作策略; , 在ComCoast试验区实施多功能防洪管理最佳方案; , 在环北海区域共享技术成果。 功能及组成 ComCoast是通过利用多条防洪线来探索海岸防洪的解决方案。相比于单一的防洪线，海岸防洪区包含一系列拥有各自功能的防洪屏障，其功能和组成引用ComCoast的主要方案表述如下: , 挡水 主要临水大堤负责抵御高海水位、海浪爬坡直至设计水位。大堤外坡(背水坡)铺设防漫顶护坡以允许较大的漫顶流量。 , 蓄水 在主坝背后的过渡区能储存一定量的漫溢海水，辅坝或高地环绕过渡区，利用沟渠或泵站进行排水。 , 洪水控制 在遭遇风暴或在正常天气情况下海岸防洪区在需要的情况下应该能进行排水，必要的排水系统有利于过渡区的水量控制，对于较大的水量可以设置泵站通过排水系统排出。如果需要，可以增加一个涵洞来增强潮汐对过渡区的影响，而且它还可以在风暴后排除多余的海水。 , 削减海浪 坝前的各种设施都可以起到削减海浪的作用。首先，浅滩可以在坝前产生较温和的海浪气候，另外早前建造的低坝、防波堤、夏季坝(辅坝)均可起到削减海浪的作用。 , 多功能区 7 过渡区可以用作多种用途，如水上运动、休闲，改善水上区域提高环境价值等，这些功能只有在经常发生规律性的洪水时才能发挥作用，当然也可以通过合理规划、布置来获得。 具体措施 以上对沿海综合性防洪策略的功能和组成的区分和描述产生以下五项主要措施: , 陆向措施 , 规律性的潮汐交换:通过结构工程，如闸门、潮汐口或管道等实现坝后区域规律性的海水 交换，从而创造海水或淡盐水生物栖息地; , 相对于现有堤防进行陆向的重新规划或布置，这包括部分或完全移除现有的堤防; , 抗溢流坝:去除大堤顶部，允许海水漫顶，并在背水坡设置护坡以抵抗溢流海水的冲刷。 漫溢的海水通过蓄或排进行有效处理。 规律性潮汐洪水 大堤重置 抗溢流坝 , 海向措施 , 前滩防护:进行滩前开发，保留或增加滩前高地或在坝前某些区域增设小堤，在遭遇风暴 时可以起到防波堤的作用; , 前滩补给:在现有沿海堤防前补充材料，包括岸线恢复和向前推移。 前滩保护 前滩恢复 前滩推进 3.3 人工土丘(或筑台) 8 居住在人工土丘上是中世纪荷兰人最传统的抵御洪水的方法，这些人工土丘的高度足以在遭遇 洪水时保持干燥。鹿特丹就是这样一个很好的例子，鹿特丹是世界上最大的港口，而荷兰及比利时 沿Westscheldt海岸又被认为是世界第二大港口，所以此地区的投资非常大。但是，对于工业资产 来说他们根本不能接受任何一点的洪水风险。炼油厂、储油区、核电站、化工厂、集装箱码头必须 确保绝对地安全。如果发生最黑暗的事情，或者巨大的风暴超过大堤的安全标准(鹿特丹附近为 10,000年一遇，其它地区为4000年一遇)，甚至大堤决口，那么坝后的地区将被几米深的洪水淹没。 洪水造成的破坏将需要几个月的时间进行恢复而且损失将是惊人的巨大。然而，位于人工土丘上的 工业设施在遭遇同样恶劣的情况下只是在几小时的高潮位时遭遇深度几厘米的洪水。 这些工业区的做法实际上部分返回了可能是最安全的防洪策略:大规模的人工土丘。为什么在 城市防洪规划中不采用这样的措施呢, 鹿特丹位于人工筑台上的炼油厂 3.4 减少洪水损失 人们一般喜欢贴近水边居住，但是这也牺牲了一些原本被用来蓄水的一些区域。Living in close proximity to water is attractive but has come at the price of land that could have been allocated to water. The possibilities of living near water are good, as long as the demands for safety and water storage are taken into consideration, now and in the future. With the increase of population and development of society and economy, more land and space are claimed for industry, housing and recreation. The area which has relatively high risk of flooding should keep preparation to live with flood by taking some individual measures, although the dike system has keep a certain safety level. Measures for decreasing flood damage 9 Inhabitants of high-risk area can take precautions to protect their homes and property and prevent a great deal of damage. In addition to the efforts extended by different level of government citizens themselves must protect their property or even take into account the heightened risk of floods in the design of buildings. Examples given include: , Raising the elevation of the ground floor: building the house on the pile or on a heightened foundation according to the suggestion of the local flood protection department; , Installing indoor heating, power and telecommunications systems as high as possible; , Use of water-resistant building materials; , Making cellars water proof. 10 Recent development in flood risk management in the Rhine Delta and coastal zone in the Netherlands Huang Bo Shandong Yellow River Reconnaissance, Design and Research Institute, Jinan Abstract: After the flooding disaster of storm surge in 1953 and the flood of Rhine in 1995, large-scale dike construction and reinforcement were carried out along the coast and the rivers in the Netherlands. The safety standard of flood protection has been raised from 1/1250 in eastern Netherlands to 1/4000 and eventually 1/10 000 along the coast. Whereas with the uncertainty of climate change in the future, the discussions on The Netherlands' flood control strategy were intensified and extended in long term consideration. Recent estimates of the change in the discharge regime of the Rhine River forecast an 3increase in the so-called design discharge from 15,000 m/s in the 1990s towards 16,800 (minimum 3scenario) to 18,000 m/s (maximum scenario) by 2100. Additionally, in the downstream deltaic area, sea level rise may hamper the discharge. For The Netherlands, sea level rise is currently estimated as between 0.2 and 1.1 m above present. Finally, in deltas and alluvial plains both shrinkage and oxidation of extensive peat layers cause the subsidence of large areas. Traditional measures such as dike heightening and strengthening are efficient in flood protection, but with social developments in the future, more claims will probably be made on space for functions other than safety against inundation, such as the landscape, the environment, ecology, inhabitation, industry, etc. Because of this, traditional dike strengthening will come up against greater social objections. This calls for a change of strategy in flood risk management of both integrated river management and spatial planning in coastal zone. 1. Background The storm surge of February 1953 caused one of the biggest natural disasters in the history of the Netherlands. The death toll of 1,853, an unprecedented high figure for Dutch standards, made a profound impression and roused emotions. The water levels that occurred during this storm surge at Hoek van Holland are the standard for the northern part of the delta area, and showed a highest water level of 3.85 m+NAP, which is 0.57 meters higher than the highest recorded water level of 3.28 m+NAP in 1894. After the flooding disaster, large-scale dike reinforcement started. The so-called Delta Project included the closure of coastal inlets and estuaries. Along whole coast and in the region of the downstream rivers the flood protection structures were strengthened in an unprecedented way. For the first time the level of protection in a certain region was normative. For Central-Netherlands the flood protection structures had to give a protection against a storm surge from the sea with a chance of occurrence of one in 10,000 per year. In 1993 and 1995, The Netherlands experienced periods of uncertainty when the Rhine River reached 3very high levels. The discharge is up to 12,000 m/s at Lobith. In the Rhine River, only in 1926 had a 3higher discharge (12,600 m/s) been recorded. Over 240,000 people were evacuated from a number of polders when the stability of the dikes seemed no longer guaranteed. Fortunately the dikes did not fail. Since 1995 structural measures have been taken. An emergency act was passed in parliament, which was called Delta Plan Great Rivers. In this act it was laid down that the entire dike reinforcement programme along the rivers had to be completed before 2000. Based on this at the dikes in the river region were indeed strengthened very quickly at the previously prescribed chance of occurrence of one in 1,250 per year. Now they protect the river region against high water levels belonging to a discharge of the Rhine 3of 15,000 m/s at Lobith. After the flood of Rhine in 1995, the discussions on The Netherlands' flood control strategy were intensified and extended to include climatic change as an additional relevant factor for the long term. Recent estimates of the change in the discharge regime of the Rhine River forecast an increase in the 3so-called design discharge (a peak discharge with a probability of 1/1250 years) from 15,000 m/s in the 31990s towards 16,800 (minimum scenario) to 18,000 m/s (maximum scenario) by 2100. Additionally, in the downstream deltaic area, sea level rise may hamper the discharge. For The Netherlands, sea level rise is currently estimated as between 0.2 and 1.1 m above present. Finally, in deltas and alluvial plains both shrinkage and oxidation of extensive peat layers cause the subsidence of large areas, a process which in The Netherlands is enhanced and maintained by a history of over 1000 years of drainage. Summarizing, the design water levels in the rivers will probably rise because of higher peak 11 discharges and a higher sea level, whereas the vulnerability of the protected areas will increase through population growth and economic development, aggravated by land subsidence. This calls for a change of strategy in the policy fields of both integrated river management and physical planning. In this paper, we first briefly introduce the present situation along the Rhine River branches and coast, and then have an analysis for the problems existed, at last introduce some solutions in flood risk management. 2. Problems on flood risk management 2.1 Discharge increase of Rhine River Due to global warming, changes will occur in the precipitation pattern in the Rhine basin area. It is expected that the Rhine, at present a combined rain and melt-water river, will increasingly become a rain river with high discharges in the winter and low discharges in the summer. The increasing winter precipitation will affect an increase in the discharge of the Rhine in winter. Summer discharge will decrease as a result of a reduced amount of melt-water and a strong increase in evaporation, the latter outweighing the effect of the smaller increase in the average rainfall in the summer. The figure bellow shows that in all of the climate scenarios (low, medium, high and dry) the expected winter discharge of the Rhine will increase even further and the summer discharge will decrease even further relative to the present discharge. Expected river discharge with four climate scenarios Increase of design discharge in 2050 The design discharge, which is used as the basis for the legal safety standards in the Flood Defences Act, is indicative for extreme discharges. In the Rhine/Meuse area, the design discharge is based on a discharge quantity, which occurs on average once every 1250 years. In legal terms, this is the maximum 12 quantity of water the river must be able to discharge without the hinterland becoming flooded. The dykes, flood plains, main channel and other related factors are dimensioned on this discharge. Following the high water levels in the Rhine/Meuse system in 1993 and 1995 the design discharge 33for the Rhine was adjusted from 15,000 m/s to 16,000 m/s. The direct consequence of this adjustment is that the current situation in the Rhine/Meuse area no longer satisfies the legal safety standards. According to the climate scenarios, in 2050 the design discharges of both the Rhine and the Meuse will have increased: the Rhine by 3–10% and the Meuse by 5–20%. This means that additional measures in the Rhine/Meuse area will be necessary to ensure that the legal standard is still met (figure above). 2.2 Sea level rise The expectation for 2100 is that climate change will lead to a further rise in sea level at the Dutch coast in the order of 20–110 cm relative to ground level. This prediction is based on an average land subsidence of 10 cm per century and takes into account both land subsidence as a consequence of time lag effects from the last ice age – to which the NAP is also subjected – and the average value for fall in the ground level due to the settling of clay and peat. However, considerable local differences in ground subsidence can occur. Note that there is a large time lag between the warming up of the oceans and the temperature rise in the atmosphere. This implies that if the average temperature of the atmosphere would be limited as a result of reduced emissions, an effect on sea level would only be realized after many centuries. One consequence of the expected rise in sea level is the need for more and larger volumes of sand to be added as beach nourishment to the coastal system to compensate for the losses of sand that now occur and to maintain current safety levels. The addition of sand to the coastal system will also ensure that the coast, estuaries and the Wadden Sea keep pace with the rise in sea level. In the future, stronger and wider dykes will be needed to offset the greater pressures arising from the rise in sea level. Over the next 50 years, the additional costs for coastal management are expected to be no more than 0.13% of the gross national product. In the case of a further rise in sea level after 2050, the costs for coastal management could increase to (much) more than the present spending level. Next to the influence of sea level rise, the occurrence of flood levels is strongly determined by the occurrence of storms on the North Sea. However, it is not yet clear how the frequency and intensity of storms will change in the future. 2.3 Increase of land subsidence Independent of the changing climate and the isostatic land subsidence, the Netherlands is also confronted with a subsidence in peat areas. Since the Middle Ages as much as 2–3 m of land subsidence has occurred in some peat areas. This land subsidence is correlated with the drainage of the peat; as a result the peat shrinks and oxidizes, disappearing as carbon dioxide (CO2) into the atmosphere. Depending on the water level, this land subsidence can be up to 1 cm per year. At this rate, a subsidence of 0.5 m will occur in some of the peat areas until 2050. If land subsidence continues, areas with thick layers of peat – in particular in the western parts of the Netherlands, where there are local depositions of peat up to 12 m thick – could, over the long term, be subjected to increased flood effects, increased surface water salinity and a water system that is increasingly difficult to manage. In several areas (for example, around Slochteren in the north-eastern part of the country) land subsidence also occurs as a consequence of gas extraction. In these areas an extra subsidence of about 60 cm is expected by 2050. The rising temperature, the longer summer season and a greater difference between wet and dry conditions (oxidation pump) will most likely result in a faster oxidation of peat. This in turn may lead to accelerated subsidence. The rates of land subsidence are the same everywhere due to variation in peat soils and differences in water management. For example, agriculture requires a relatively deep drainage level, whereas urban areas in peat areas require a relatively high water table in order to prevent wooden pile foundations from decaying. Land subsidence in the peat areas therefore leads to an increasingly fragmented water management system, to a stronger salt seepage (detrimental for agriculture) and to damage resulting from the subsidence of roads and buildings. Various provinces – especially in the western part of the Netherlands – have consequently included measures to counteract land subsidence. These measures, however, only affect 4% of the total peat land area. 13 Land subsidence in peat land areas 2.4 Storm surge hazard Since 1962 the number of storms per year has decreased. The figure bellow shows the distribution of the 700 most extreme storms in the Netherlands over the past 41 years. The wind speed associated with these storms was, depending on the location within the country, more than 11–16 m/s; this is equivalent to a wind force of 6–7 on the Beaufort scale. Moreover, even if only the 300 or 500 most exceptional events are considered (heavier storms), the picture does not change: the number of storms in the Netherlands is decreasing. To what extent this decrease is correlated with rising temperatures is not clear. Distribution of the 700 extreme storms in the Netherlands over the past 41 yearss The large uncertainty in the effect of climate change on the storm patterns in the Netherlands means that an understanding of changes in the likelihood of storm floods is far from complete. Recent research with large-scale models indicates a possibility of ‘super storms’ occurring within the orders, with the chance of significantly higher wind speeds than the Netherlands have experienced in the 20th century. Further research with more refined models is necessary in order to understand the underlying processes. 3. Measures on flood risk management 3.1 Room for river Storage of water along the Rhine Branches - detention measures By detention it means that a segment of the discharge peak is shaved and is temporarily stored in a diked-in area. Once the worst has passed, the temporarily stored water is released again and discharged. 14 Detention limits the quantity of water to be discharged for the section located downstream of one (or more than one) of the Rhine Branches. This means that detention must occur upstream if it is to fulfill its purpose. For The Netherlands, this translates into as close to Lobith as possible. This may be understood as follows: the difference between the current design discharge of 15,000 33and the expected discharge of 16,000 m/s is 1,000 m/s. In order to prevent dike raising downstream 3exclusively through the use of detention methods, this amount of 1,000 m/s must be allowed in via intakes into one or more detention areas. The total storage capacity necessary is determined by the difference in the height of both tops in relation to the shape (and especially the duration) of a flood wave. In this way, it may be calculated how much water must temporarily be stored within one or more dike rings. 33For the 1,000 m/s mentioned here, a storage capacity of some 170 to 200 million m is required for a flood wave lasting several days and having an ‘average shape.’ At a depth of 5 metres, this means a necessary surface area of some 3,500 to 4,000 hectares. Similar depths are conceivable in the relatively low-lying areas within the dikes along the Boven-Rijn and the Waal, but along the Neder-Rijn and IJssel much shallower water depths would have to be realised. The higher the storage areas lie, the less the water depth will be and proportionately larger the surface area will be. Discharge of water via the Rhine Branches In comparison with storage, the most important difference is with storage, the amount of discharge is reduced; and with measures to increase the discharge capacity, only the water level is lowered while the discharge remains the same. Storage, provided it is carried out for long enough, is beneficial for the region downstream of where the measure is being implemented and for only a very short distance upstream. Increasing the discharge capacity is advantageous for the section of river upstream from where a measure is being executed. The reason for this is that the backwater effect of a narrowing, another type of obstruction, or a substantial ‘hydraulic roughness’ is diminished, or that the cross-sectional profile is increased. The creation and implementation of measures which increase the discharge capacity must be first carried out downstream, with subsequent measures carried out in an upstream direction. There are many measures to achieve the purpose, we can divide these up into three large groups: , measures in the low flow channel; , measures in the flood plains; and , measures in the areas protected by the dikes (setting back dikes, etc.) Within each group there are then specific types of measures as following. , Lowering of low flow channel From the calculation it appeared that dredging of the low flow channel bed can produce a water level reduction between 20 and 30 cm over a distance of some 50 km. 15 , Lowering of groynes By reducing the height of groynes, more water flows over them and less flows through the low flow channel. This can result in a decrease in erosion and a delay in the occurrence of erosion in the low flow channel. The contribution of groyne height reduction to counteracting unwanted erosion is then considered a favourable incidental circumstance. , Flood plain excavation Flood plain excavation is a measure by which the gradual development of heightening by sedimentation on flood plains may be counteracted. It may be combined with clay mining, as was recently applied within the framework of the most recent dike reinforcements, and/or with nature development. , Removal of hydraulic bottle necks Based on the water level slope line in all three Rhine Branches, it was first determined where the hydraulic bottlenecks were to be found. Using this method, 254 bottlenecks were found. Next, with the aid of topographic maps, these were identified as flood-free areas, bridge abutments, ferry ramps, etc. Some of these bottlenecks comprise a considerable surface area, such as floodfree (factory) areas. Ferry ramps, bridge abutments and summer embankments are, conversely, much smaller. The water-level lowering effect of the measures varies so much that the efficiency also differs greatly. 16 At that point, two criteria were applied to select 60 measures (including 18 small-scale dike relocations) that were taken into consideration in the further study, namely: , a water-level reduction effect of at least 1 cm (‘otherwise it’s not worth the effort’); and , an efficiency grade of at least 2 mm/million Euro (‘otherwise it is a relatively expensive measure’). , Large-scale setting back of dikes The options for setting back dikes are rather expensive, but sometimes also extremely effective. Despite the costs which vary from 4 to 50 million Euro, the setting back of dikes all surpass the efficiency level of 2 mm water level reduction per 1 million Euro. The achieved water level reduction of individual setting back of dikes can run into several tens of centimetres, but in contrast there are measures in this category that only produce several centimetres worth of reduction; nevertheless, these also satisfy the criteria that were formulated earlier as the minimum for measures at hydraulic bottlenecks. Setting back dikes is particularly effective in cases of real narrowing of the flood plain which cause backwater effects quite a distance upstream. This is why the consequences of setting back dikes for the water level also continue to work relatively far upstream. A local reduction of some 10 to 20 cm is realisable per measure. Along the Waal and the Neder-Rijn/Lek, all of the settings back of dikes together can result in a maximum reduction of 60 cm. , Green river for urban bottlenecks River widening and deepening measures upstream of such a narrowing do not offer a solution, since widening and deepening as it were occurs in or around a ‘reservoir,’ without the valve which empties the ‘lake’ being opened. Measures downstream of such a narrowing have little effect on the water levels since this would be comparable to drawing water using a closed valve. And even this does not help. This means that the urban bottlenecks do harm the effectiveness of the other measures and may work unfavourably in terms of the costs of combined alternative strategies. It is obvious that the only thing that can offer relief for these urban bottlenecks are measures in the dike-protected areas; after all, flood plains are (almost) entirely lacking. As has been mentioned, similar measures were not taken into consideration in an earlier stage in connection with the expected futility of administrative feasibility, seeing as the measures involved would have to be executed in a relatively densely built-up area. For this reason, the green rivers have been created for measures involving urban bottlenecks. Green rivers are in fact flood plains between two (guiding) dikes where water does not flow through during low discharges, but does flow during floods. They may be used for agricultural purposes or may be designed for nature and/or recreational areas: they are, in short, ‘green.’ 17 3.2 Combined functions in Coastal defence zones – “ComCoast” Concept and aims In the coming years climate change will increase the physical loads on coastal defences all over the world. Traditionally the Netherlands has protected it selves against the growing threat of flooding, by heightening the dikes. However, with the continuing sea level rise, it becomes more and more evident to find alternative and innovative strategies, without just heightening the dikes. ComCoast develops such flood risk management strategies, with gradual transitions from sea to land, in order to create integrated defensive zones including wider environmental functions, such as recreation, fisheries, tourism and nature creation. The ComCoast concept aims to create multifunctional flood management schemes with a more gradual transition from sea to land, which benefits the wider coastal community and environment whilst offering economically sound options. The aim of ComCoast is as following: , to explore the spatial potentials for coastal defence strategies for current and future sites in the North Sea Interreg IIIb region , to create and apply new methodologies toevaluate multifunctional flood defence zones from an economical and social point of view. , to develop innovative technical flood defence solutions to incorporate the environment and the people and to guarantee the required safety level , to improve and apply stakeholder engagement strategies with emphasis on public participation , to apply best practice multifunctional flood management solutions to the ComCoast pilot sites , to share knowledge across the North Sea region. Technical functions and components The ComCoast project searches for alternative coastal defensive solutions using a multiple line of defence strategy. In comparison with a single line defence, a coastal defence zone has a range of components (lines) each with its own function. First these technical functions and its components are formulated from which the main ComCoast solutions can derived. , Water retaining The primary dike retains high sea levels and wave run-up, up to the design level. The inner slope can have an overtopping-resistant revetment, which permits a greater overtopping discharge. , Water storage The area behind the primary dike is a transitional area able to store the overtopping seawater. A secondary dike or higher grounds, encircles the transitional area. Or the water is handled by large ditches or pumping stations. , Water control / management During storms and in normal weather conditions the coastal defence zone should be able to drain off water when necessary. First, a drainage system facilitates water control in the transitional area. For larger quantities of water, a pump installation can be installed to support the discharge of water by the drainage system. If desired, a culvert can be added to increase tidal influence in the transitional area. A culvert can also be used to drain off excessive salt water after a storm. , Wave reduction Several elements in front of a dike yield wave reduction. First, a shallow foreshore creates a moderate wave climate in front of the dike. In addition, wave reduction can also be achieved when there is a previously constructed lower dike, a breakwater or a summer dike. , Multifunctional use of area The transitional area can be used for several purposes, for example aquatic sport, recreation, the development of aquatic areas and to enhance environmental values. This is only the case when the area is flooded regularly. This can also be obtained by Managed realignment. Comcoast solutions The distinguishing of the functions and components as discussed in the previous paragraph has led to five main ComCoast-concepts: , Landward solutions: 18 , Regulated tidal exchange is the regulated exchange of seawater to the an area behind fixed sea defences, through engineered structures such as sluices, tide gates or pipes to create saline or brackish habitats. , Managed realignment involves the placement of new Managed realignment flood defence landward of the existing flood defences. This would be achieved trough the partial or complete removal of the existing flood defences. , Overtopping resistant dike involves the replacement of the top of the dike and its inner slope with a revetment that will not wear away by severe overtopping. The overtopped sea water will be handled in the coastal zone at the landward side of the dike (drainage/storage) , Seaward Solutions: , Foreshore protection involves reclamation works to maintain or to create higher ground and in some situation small dikes in front of the primary dike, which act as breakwaters in case of a big storm , Foreshore recharge involves the placement of material in front of the existing coastal defence system. It includes to restore the coastline and to advance the line. 3.3 Live on mound Living on mound is a very traditional solution which was in use in the early Middle Ages in the Netherlands. These are artificial hills which are high enough to remain dry during floods. Rotterdam is a good example in such way. Rotterdam is the largest port in the world, and the combination of the harbors along the Westerscheldt in both the Netherlands and Belgium could be considered the second-largest harbor in the world. So, huge investments have been made in this area. But they have shown that captains of industry will not accept any risk of flooding at all. Refiners, oil terminals, nuclear plants, chemical industries, container terminals had to be entirely secure. If the darkest situation occurs, and the safety limit of the dams (one in 10,000 years near Rotterdam and 1:4000 years in other areas) is overtaken by an enormous storm, even these dams will break and the lands behind will become drowned in water meters 19 deep. The damage will take many months to repair and the losses would be amazing great. However, the industrial complexes on their artificial mounds, in the same worst case scenario, suffer a few centimeters of flooding during the few hours of high tide. The industry made a partial return to the safest possible strategy for flood prevention: large-scale artificial mounds. Why don’t use this measure in urban planning? Refiners on artificial mound in Rotterdam 3.4 Living with flood Living in close proximity to water is attractive but has come at the price of land that could have been allocated to water. The possibilities of living near water are good, as long as the demands for safety and water storage are taken into consideration, now and in the future. With the increase of population and development of society and economy, more land and space are claimed for industry, housing and recreation. The area which has relatively high risk of flooding should keep preparation to live with flood by taking some individual measures, although the dike system has keep a certain safety level. Inhabitants of high-risk area can take precautions to protect their homes and property and prevent a great deal of damage. In addition to the efforts extended by different level of government citizens themselves must protect their property or even take into account the heightened risk of floods in the design of buildings. Examples given include: , Raising the elevation of the ground floor: building the house on the pile or on a heightened foundation according to the suggestion of the local flood protection department; , Installing indoor heating, power and telecommunications systems as high as possible; , Use of water-resistant building materials; , Making cellars water proof. 20 Measures for decreasing flood damage Reference Machteld Van Boetzelaer and Bart Schultz (2006), Recent developments in flood management strategy and approaches in the Netherlands. Utrecht, the Netherlands; Delft, the Netherlands. Dick De Bruin and Bart Schultz (2006), A simple start with far-reaching consequences, The Hague, The Netherlands; Utrecht, The Netherlands. Dick De Bruin (2006), Irrigation and Drainage 55: S1-S2 (2006), Similarities and differences in the historical development of flood management in the alluvial stretches of the lower Mississippi Basin and the Rhine Basin, Published online in Wiley InterScience (www.interscienc.wiley.com). Wilfried Ten Brinke (2005), The Dutch Rhine, Veen Magazines B.V., Diemn, The Netherlands. Fundermentals on Water Defences (1998), Technical Advisory Committee on Water Defences, The Netherlands. Wim silva, Frans Klijn, Jos Dijkman (2001), Room for the Rhine Branches in The Netherlands, IRMA-SPONGE, RIZA, WL/ delft hydraulics, The Netherlands. Comcost- Innovative solution for flood protection and regional development, 2005. Rijkswaterstaat DWW, Delft, The Netherlands. www. comcoast. org. Frans Kijn, Michael van Buuren, and Sabine A.M. van Rooij (2002), AMBIO: A Journal of the Human Environment, Vol. 33, No.3, pp. 141-147. Flood-risk management strategies for an uncertain future: living with Rhine River floods in The Netherlands. stA Different Approach to Water, Water Management Policy in the 21 Century (2000), Ministry of Transport, Public Works and Water Management, The Netherlands. 21