圩田景觀：荷蘭低地的風景園林

作者：（荷蘭）斯蒂芬·奈豪斯

翻譯： 韓冰

校對： 郭巍

Text: Steffen Nijhuis (The Netherlands)

Translator: HAN Bing

Proofreading: GUO Wei

圩田景觀：荷蘭低地的風景園林

作者：（荷蘭）斯蒂芬·奈豪斯

翻譯： 韓冰

校對： 郭巍

Text: Steffen Nijhuis (The Netherlands)

Translator: HAN Bing

Proofreading: GUO Wei

荷蘭低地大部分都是圩田，人們通過人工控制水位，可以在圩田中工作和生活。人與水之間數百年來此消彼長的相互作用孕育了豐富多樣的圩田景觀。地質上底層土壤的差異、水與土的動態變化以及人類干預過程生成了種類繁多的圩田形式。本研究從風景園林的視角系統地探索了荷蘭圩田景觀形態塑造的成就，并將其視覺化呈現。圩田，不僅僅可以看作水力學現象的產物，同時也是空間結構和文化表達的結果：圩田景觀不僅僅可觀、可游，同時反映了荷蘭文化。通過探索圩田景觀形態，我們可以“解讀”荷蘭低地的場地精神，以便獲取其后隱藏的信息及設計知識，并成為下一步發展的線索。本研究分兩個階段分析了圩田景觀的風景園林化形式，并借助系統分析和制圖的方式呈現了圩田景觀的共通點和差異性。首先，調研所有圩田并進行數字化，得到第一張系統化的荷蘭圩田地圖。而地理信息系統（GIS）的運用不僅僅保證了精確度，也為使得將信息融入地圖并組建空間數據庫成為可能。在第二個階段，進一步研究圩田這一低地基本景觀單元。通過對選定的不同圩田類型（海岸黏土圩田、河流圩田、湖床圩田、和泥炭土圩田）的實例進行形態學分析，列舉并解讀每一種圩田類型的特征。截至目前，圩田景觀的研究大多基于自然和歷史地理學角度，而上述分析使得人們可以觸及圩田景觀的空間設計法則，并得到這類景觀保護和轉型的線索。

圩田景觀；設計分析；荷蘭低地；形式分析；類型學；圖術

韓冰/女/1991年/山東人/碩士生/ 北京林業大學園林學院（北京 100083）

Translator:

Han Bing was born in 1991,native of Shangdong province. She is a postgraduate of Landscape Architecture School in Beijing Forestry University.(Beijing 100083)

校對簡介：

郭巍/1976 年生/男/浙江人/ 博士/ 北京林業大學園林學院副教授/ 荷蘭代爾伏特理工大學（TUD）訪問學者/研究方向為鄉土景觀（北京 100083）

Proofreader:

GUO Wei, who was born in 1976 in Zhejiang, obtained his PHD degree in 2008. Now, he is an associate professor in School of Landscape Architecture, Beijing Forestry

University and a visiting scholar of Delft University of

Technology, the Netherlands. His major focuses on vernacular landscape.

1　導論

圩田分布于世界各處沿海和沖積形成的低地上。人們控制這些墾地的水位，并在其上工作和生活。人與水之間此消彼長的相互作用持續了約數個世紀之久，孕育出豐富的圩田景觀。然而，氣候變化和經濟發展使圩田景觀受到威脅。隨著海平面上升而日益增長的洪水威脅、大規模排水導致的持續沉降以及急速的城市化，應對這些變化并做到有規劃地發展迫在眉睫。為了保護這一珍貴文化遺產，我們需要收集相關經驗和知識，來找尋保護這種景觀和引導其轉型的線索。將圩田看作是一種水力學現象或歷史過程產物的研究有大量，但很少有研究將圩田看作一種空間結構和文化表達的結果：圩田景觀不僅僅是可觀可游的，同時也是文化的展示[1]。為了填補這一研究領域的空白，代爾夫特理工大學（荷蘭）的風景園林團隊展開了一個長期研究項目，重點從風景園林層面來研究圩田景觀。這個項目除擴充知識，還促進了對于這類具有人文意義但又水害頻發的低地景觀的關注度提升，也使人們意識到過度耕種及大量建設的現狀問題。同時，該項目也為政策制定以及如何從風景園林角度進行規劃和設計提供了參考。本研究強調圩田景觀是一種文化的表達——而不是單純水利工程的產物——并借助圖術、案例研究以及比較研究的方式，推斷和解讀這些“圩田景觀（polderscapes）”的異同點。研究主要成果有《海的土地》（2007），其研究重點為湖床圩田；《荷蘭圩田地圖集》（2009），從更寬泛的角度敘述圩田；以及《綜合性荷蘭圩田地圖冊》（2013），是針對荷蘭圩田的首次系統性綜述[2]。此外還有近來醞釀的以國際視角研究圩田的《世界圩田景觀》。

本文旨在從風景園林角度，概述荷蘭圩田景觀的研究。這項研究理清了圩田背景的發展脈絡，系統研究并描繪了多種尺度下的圩田景觀豐富性，完成了基于GIS的覆蓋國土面積的地圖繪制，并在代表案例研究的基礎上，對不同圩田類型的特征進行了描述。文章的開頭概述了作為一種文化產物的荷蘭圩田景觀。接下來則是研究策略的介紹，詳盡說明了如何運用形態研究及地圖作為可視化思維和交流的工具。因此，本文在討論荷蘭圩田地圖的繪制過程時，重點敘述了其主要信息來源以及圩田單元的識別。接下來，則重點說明在研究單個圩田特征時采用的分析框架，并列舉荷蘭一處17世紀的湖床圩田貝姆斯特爾圩田（Beemster）為例。最后，在文章的末尾給出結論。

1 歐洲西北部的荷蘭低地（S. Nijhuis提供，代爾夫特理工大學）The Dutch lowlands in Northwest Europe (S. Nijhuis, TU Delft)

2 作為文化產物的荷蘭圩田景觀

圩田景觀可以被定義為：“一處平坦的區域，最初面臨高水位的威脅（來自地表水或地下水的永久或季節性的威脅）。在圍墾后，這一區域能夠獨立調控水位，并由此與其原來所處的水文狀況隔離開來[3]。因而，衍生的圩田景觀是一種文化和歷史表達結果，因當地自然環境、知識技術與管理之間的相互作用而產生?！?這個概念涵蓋了門類紛繁的圩田景觀，包括長期處于積水環境下的泥塘沼澤、淺海和湖床地區，以及暫時受到水澇和積水影響的淺潮低地及沖積平原?；谏鲜鏊h境的區別，可以總結出3類圩田景觀：（1）低洼地圍墾產生的圩田景觀，（2）開墾近海產生的圩田景觀，及（3）湖泊排水后產生的圩田景觀[4]。盡管這些圩田景觀有一定范圍的一致性，但由于地質上底層土壤、水與土的動態變化以及人類干預過程的差異，因而在世界范圍內，圩田的形式極為多變。雖然自然條件是形成差異的重要因素，但人類在圩田發展過程中也起到決定性的作用。這是因為景觀是人類與自然因素相互作用的結果。正如地理學家卡爾·索爾（Carl Sauer）所說：“文化是催化劑，自然是媒介，形成的文化景觀就是結果?！盵5]

2 作為花園的荷蘭：Joan Blaeu繪制的十七世紀地圖，顯示出了荷蘭的北荷蘭省幾何式的圩田結構（代爾夫特理工大學圖書館，Tresor提供）Holland as Garden: seventeenth century map by Joan Blaeu showing the geometrical polder landscape of North Holland, the Netherlands (TU Delft Library, Tresor)

3作為圩田的花園：奧蘭治王室的Honselaarsdijk王室花園成為了荷蘭園林的代表，見Abraham Bega和Abraham Blooteling于1683年描繪的示意圖。（皇家檔案館提供）The garden as polder: the princely garden of Honselaarsdijk became a representation of the Garden of Holland with the House of Orange as its protector as depicted by Abraham Bega and Abraham Blooteling, 1683 (Royal Archive)

2.1土地的測量、開墾及劃分

在荷蘭的萊茵河、默茲河及斯凱爾特河的三角洲地帶（圖1），生產性圩田景觀有十分悠久的規劃、建造及維護的歷史。圩田景觀源遠流長，與荷蘭文化融合共生。從自然角度上來看，這一點也毋庸置疑，因為圩田占據荷蘭國土面積的一半，另從生產知識和水資源管理角度來看也解釋得通。諸如大堤、水壩和涵洞等最古老的水利基礎設施，最早可以追溯到羅馬之前的時代。但直至公元1世紀羅馬人遷入后，才開始大規模泥炭沼澤的開墾[6]。隨后于10世紀開始了對荷蘭其余的泥炭沼澤進行系統性開墾。16世紀的工程、測量及制圖領域的發展也為土地開墾注入了新的動力。堤壩建設的進步（約1576年）以及風車排水的運用（自1533年開始）使得圍湖造田成為可能。同樣，16世紀30年代發明了三角測量法，土地測繪師開始使用測量鏈和羅盤進行更大范圍區域的測量，誤差也減少了[7]。測量師開始繪制地圖，用于土地清算、分割（設計布局）、對地塊進行管理及優化等，因而他們在堤壩建設、水利施工以及大型土地開墾項目中起到至關重要的角色。隨著資本主義的興起，土地在當時成為一種商品，因此地圖也成為具有法律效應的工具。地圖是植根于公民美德與區域開墾的社會面貌的表征[8]。

發展建設和耕作的需求潛力使湖床圩田的實現在經濟上成為可能。富有的商人不斷投資大型項目，開墾出肥沃土地將糧食供給城市，并建設遠離城市污染的舒適的居住區域。湖泊轉化和開墾成可盈利的耕地，在其上建造美麗的莊園和賺錢的牧場。截至17世紀40年代，阿姆斯特丹土地面積相比16世紀增長了40%。這要歸功于貝姆斯特爾（Beemster）（1608-1612）、普梅爾（Purmer）（1662）、沃爾默（Wormer）（1625）以及斯荷爾姆（Schermer）（1631）四處湖床圩田的開墾。圩田布局由測量師決定，他們都接受過土地測量和制圖繪圖的規范訓練。圩田布局并非是單一的設計結果，而是結合了土地分割的實用性。土地被分割成為不同尺度的矩形塊，與各處的土壤組成、水利工程設施及土地利用最優化等相適應，最終形成多層格網結構。借鑒幾個世紀以來累積的經驗，測量員能夠輕松地應用靈活多變的方格來處理場地環境[9]。西蒙·斯蒂文（Simon Stevin）（1548-1620）是荷蘭城市規劃、聚落和防御工事建設的奠基人。他深諳這些原則，并將這種靈活而世俗化的規劃實踐總結成文。其中多層格網的對稱性、軸線組織方式以及其靈活的矩形模數體系也可以應用在城市規劃之中[10]。

2.2荷蘭作為花園

土地開墾、耕作和創造這種典型的幾何景觀成為了荷蘭造園藝術發展的基礎，以至于國家本身也被認為是一個花園而人民則是園?。汉商m作為花園或者Hortus Batavus（圖2）[11]。由17世紀土地開墾和測量技術形成的線性模式將開墾的土地轉化為真正意義上的文化景觀，人們在其中出于經營獲利和美化目的而建造了鄉村田莊[12]。這些田莊附屬花園的幾何秩序完美融合了圩田景觀的規則線性[13]。其中一例是腓特烈·亨利（Frederik Hendrik）（1584-1647）在洪塞拉爾戴克（Honselaarsdijk）的皇家花園。該花園后來成為了荷蘭式園林的代表，歸屬奧蘭治王室（the House of Orange）（圖3）[14]。與其他許多田莊類似，洪塞拉爾戴克融入了周邊的圩田景觀，局部與特有地形條件吻合，并借鑒了土地開墾及分割的常用手段。因而，它是實用性和舒適感的融合，象征了人類對自然的優化和改造：這個花園就是一處圩田，并且是荷蘭作為花園這一壯闊環境中的一部分。

16及17世紀是荷蘭文化形態形成的關鍵時期。圍墾過程作為一種政治、經濟和制圖共同作用下的文化產物形式。圩田景觀的規則化布局與現代觀念中關乎文明、所有權和治國方略的方面不謀而合。盡管“圩田（polder）”這一術語起源于荷蘭（17世紀：“polre”），但這并不意味是荷蘭人最先開始營造類似景觀的。自公元前3 000多年前的人類文明曙光開始，在美索不達米亞地區，人們已懂得利用水土環境營造生產性景觀，并修建了用水基礎設施，來抵御洪水（堤壩及水壩）并管理用水（排水或灌溉）[15]。然而，在16和17世紀，荷蘭形成了開墾圩田的理念。圩田成為一種國家象征，體現了荷蘭人在努力馴服水土自然時的堅強意志和技術特長。同時，墾田行動也似乎成為了一種荷蘭出口產品，影響了從12世紀直到目前歐洲范圍內甚至世界其他區域的圩田營造[16]。

今天的荷蘭圩田景觀正經歷向多功能空間的轉型。除了農業屬性外，人們越來越多地看重圩田景觀的休閑觀光、自然保護與蓄水、房屋建設等功能。這些發展為維持空間質量及鄉村景觀特性帶來了壓力。而解決這些空間問題的關鍵，則隱含在圩田景觀自身所蘊含的豐富形式之中。

3　理解圩田景觀

4 地圖在知識積累和設計形成循環中起到了推動和媒介作用：一個從數據到知識經驗，從知識經驗引導創新的自我循環過程（S. Nijhuis提供，代爾夫特理工大學）Maps as facilitator and mediator in the knowledge formation-cycle and design generation-cycle: an iterative process from data to knowledge, from knowledge to invention (S. Nijhuis, TU Delft)

形態學研究被用于荷蘭圩田景觀的研究中。通過探尋圩田景觀蘊含的形態學，可以“解讀出”荷蘭低地的場地精神，由此可以獲取其中隱含的信息和設計要點，作為探究未來發展的線索。事實證明形態研究，或稱形式研究是有意義的，因外在形態可作為解讀和分析景觀的媒介。自然、生物和社會文化的作用力因其效果明顯，也成為研究的重點所在。形式分析屬于一種解釋性研究手段（也可以稱其為：闡釋學）。通過形式分析及對景觀可視化表達形式的解讀來獲取相關信息。我們可將形式看作一個純粹的沒有預先設定意義的解釋系統，因而根據假說推斷結論的程度便降至最低[17]。這一類研究的目的不僅僅是為了了解和研究景觀建筑，同時也側重于探索其空間智慧，提升理解和設計其空間組成的能力，以及找尋不同尺度之間的聯系。形態學研究有三個特征[18]。首先，包含所有尺度的景觀，從花園尺度到區域性尺度。第二，將景觀結構的立體特征與其空間及動態的模式結合起來。第三，形態學研究的方法關注這種形態是如何形成的。因為景觀是時間的產物——時間上經歷了概念的形成、發展以及突變。此外，形態學研究也同樣應用于在城市設計、建筑、藝術與地理學領域[19]。

在景觀形態的研究中，地圖被看做發掘信息的基礎性工具。幾個世紀以來，人們借助地圖這一工具描繪和了解景觀。地圖幫助人們理解人類社會或自然世界中的一些事物、概念、情形、過程或事件[20]。地圖中蘊含了大量信息，且十分具有說服力，因而是獲取知識的重要手段。這種手段應該與被定義為圖術（Mapping）的手法區分來，圖術是通過整理空間信息而獲取知識的手段。它是一個過程，而不是一項既成成品。圖術需要制圖探索活動，利用地圖圖紙本身以及地圖繪制作為中介來視覺化建構和傳達空間信息。因而可以說，圖術并不完全與現存環境一一對應，故需要對其包含的信息進行解讀和重構。圖術需要通過理性和系統的方法“消化”信息，因而這是一個個人過程，受到個體選擇和判斷或稱視覺思維的影響。與此同時，視覺的呈現有助于信息傳播，表達出其中的關系、結構和模式，亦即視覺傳達。對地圖的詳細研究、比較研究以及地圖疊加分析都是有用的分析方法[21]。這些探索有助于風景園林研究人員獲得其中未知或潛在的信息，是獲取空間知識并指導設計過程的基礎。研究地圖同樣有助于數據的解讀，從中可以總結出知識經驗，啟發設計靈感（圖4）。

5 綜合性荷蘭圩田地圖（S. Nijhuis 及M.T. Pouderoijen提供，代爾夫特理工大學）Comprehensive polder map of the Netherlands (S. Nijhuis & M.T. Pouderoijen, TU Delft)

6 早期圩田地圖的案例，由Floris Balthasars在1611年繪制，范圍是代爾夫特及其周邊地區（國家檔案館提供）Example of an early polder map in the surroundings of Delft made by Floris Balthasars, 1611 (National Archive)

研究者分兩步對圩田景觀的形態進行分析，通過系統分析和制圖來揭示其異同。首先，對所有圩田進行測量，將成果數字化，獲得第一張綜合性的荷蘭圩田地圖。第二，將圩田看作基本景觀單元進行進一步研究。

4　綜合性的荷蘭圩田地圖

7 地形軍事地圖1850-1864（TMK）的細部（代爾夫特理工大學圖書館，Tresor提供）Detail of the Topographic and Military Map 1850-1864 (TMK) (TU Delft Library, Tresor)

8 水利地圖第一版1865-1891的細部（代爾夫特理工大學圖書館，Tresor提供）Detail of the Hydraulic Map first edition 1865-1891 (TU Delft Library, Tresor)

9高度精確的現代荷蘭高程圖（AHN5）細部（M.T.Pouderoijen提供，代爾夫特理工大學圖書館）Detail of the highly accurate Height map of the Netherlands, AHN5 (M.T. Pouderoijen, TU Delft)

10 圩田景觀的系統性描述過程：（1）對水管理地圖進行數字化/地理信息校準處理，（2）水管理單元矢量化，（3）邊界的確定，（4）對地形及軍事地圖數字化/地理信息校準處理，（5）堤壩的矢量圖示化，（6）評估與整合（S. Nijhuis 及 M.T. Pouderoijen提供，代爾夫特理工大學）Systematic delineation of the polder landscape: (1) digitized/geo-rectified water management map, (2) vectorization water management units, (3) definition boundaries, (4) digitized/geo-rectified topographic & military map, (5) vectorization dikes, (6) evaluation and integration (S. Nijhuis & M.T. Pouderoijen, TU Delft)

荷蘭的圩田地圖系統地呈現了豐富多樣的圩田景觀形式（圖5）。從某種意義上說，它的獨特性主要在可將整個荷蘭圩田景觀包括在內，且對圩田這種水利和景觀實體再加以突出。盡管荷蘭圩田制圖學歷史悠久，但水委會（waterboards）留存的地圖多描繪單個圩田或圩田的一塊局部區域，并沒有全國范圍的圩田地圖（圖6）。書籍中經常出現對圩田景觀的描述，但主要是從水管理和歷史意義的角度。1884年A.A. Beekman出版了《圩田上的荷蘭》。本書為荷蘭圩田的概念描述打下了科學基礎。后繼者追隨他的腳步，眼光更多投向了特定區域圩田或湖床圩田。這張綜合性的荷蘭圩田地圖首次嘗試繪制整個國土范圍的圩田，并將圩田作為一種水利營建和景觀結構。深刻理解圩田這一屬性，對于解讀這種獨特的空間形態來說是必要的。而在覆蓋全國的這份圩田地圖中可以找到這一聯系，因而可以說圩田地圖是帶領荷蘭圩田制圖方面向前邁進的一步[22]。

自20世紀伊始，由于爆發式的城市化和經濟發展，荷蘭景觀發生了很大改變，因而幾乎不可能通過當代的地形地圖來辨認出圩田單元，尤其是歷史較長的那些。為了解決這一問題，研究者重點考慮將現存完好的景觀與可獲取的詳細地圖資料結合起來。收集的19世紀至20世紀間的歷史地圖為拼湊一張完整的圩田地圖提供了可能。這一時期的城市——也有例外的情況——外圍的古代防御工事依舊完好。景觀則形成于先前的幾個世紀，且其狀態大多數情況下并沒有改變。19世紀的測量術已經有很高的精確度，因而那個年代的地圖圖像投射在現在的地圖上，也不會有太多出入。這一時期同樣開始了覆蓋全國的1850定位。地圖經過掃描后，轉存為像素構成的數字文件。每一張數字地圖都運用了地理信息系統（GIS）手段進行了地理修正和地理坐標定位。由此，對歷史地圖的平面進行了調整，并參考荷蘭的標準坐標系系統進行擴充[23]。這使得地圖得以精確地合并轉換為地理坐標信息。

下一步則是對地圖的解讀。將圩田識別為空間水利景觀單元，并將其矢量化（圖10）。參考荷蘭高程圖評估和修正所繪制的地圖，同時將1850年后開墾的地塊鑒別出來。得到的成果是一張基于GIS的荷蘭圩田地圖，展示了尺度在1:10 000至1:25 000不等約9 087個圩田單位，其中包括了圩田的各種類型：低地圍墾、圍海以及墾湖得到的圩田。更具體地說，這張地圖展示了水利單元及其明確的邊界。圩田在這一框架結構中作為空間單元存在。

GIS的運用不僅保證了工作的精確度，同樣也使地圖鏈接信息成為可能，地圖轉變為空間數據庫，包含了：名稱、開墾年代、數據資料等等。研究人員參考荷蘭地理字典（Geographical Dictionary of the Netherlands）（Van der Aa，1839-1851，荷蘭語），創立了一處字典中提及的所有圩田的數據庫（圖11）。這些信息恰好可以與地形軍事地圖-（TMK）和水利地圖中的數據相互補充及核對?？梢赃\用GIS空間化這些信息，進行分析和視覺化。同樣，它可以為研究圩田景觀隨時間的發展提供支持（圖12），也可用于對圩田特性的計算分析如：圩田尺度、形式復雜度及圩田形態。1864地形軍事地圖（TMK）以及1865-1891水利地圖第一版的繪制，兩張圖紙的比例都是1:50 000（圖7-8）。這兩張類似的歷史地圖成為了繪制圩田地圖的基礎。接下來，借助機載激光雷達點云數據（每平方米約8個點位）繪制出高度精確的現代荷蘭高程圖（AHN5），這樣可以確保范圍的精確并借此識別新生圩田（1850年以后開墾的圩田）（圖9）。

11 圩田數據庫截圖（S. Nijhuis提供，代爾夫特理工大學）Screenshot of the polder database (S. Nijhuis, TU Delft)

繪制圩田地圖的第一步是對1850-1864地形軍事地圖（TMK）以及1865-1891第一版水利地圖, 共計366張地圖進行數字化及地理坐標

為了把握圩田形式多樣性的關鍵，研究人員基于其自然地理特性來解讀圩田地圖。不同的特性來自于基質（巖石及地形）、土壤和水力學方面的差異。人類活動的影響也是其中的一個因素。因此，在非生物因素和人類活動間的相互作用下，結果會出現有代表性的幾種特性的特定組合。因此形成了基于自然地理景觀的圩田類型學，每一種類型都有許多突出的形態特征（圖13）。圩田的類型包括：海洋黏土圩田（例如Tzummerpolder）、泥炭土圩田（例如Ronde Hoep）、河流圩田（例如Alblasserwaard）以及艾塞爾湖周邊圩田（例如Noordoostpolder）。因為湖床圩田（例如Beemster貝姆斯特爾圩田）同樣是從泥炭土湖泊開墾形成的，因此可以看做泥炭土圩田的一個子類型。

5　圩田語法的識別

12 圩田地圖細部，展示了西南部圩田景觀的發展過程（M.T. Pouderoijen提供，代爾夫特理工大學）Detail of the polder map showing the development of the polder landscape in southwest (M.T. Pouderoijen, TU Delft)

13 基于自然地理景觀的幾種圩田類型（S. Nijhuis提供，代爾夫特理工大學）Polder typology based on physical-geographical landscapes (S. Nijhuis, TU Delft)

14 選取的圩田是荷蘭種類繁多圩田景觀的代表（從左到右）： Noordoostpolder，Tzummerpolder，Polder Biesbosch，Kockengen（Paul Paris攝）The selected polders represent the wide variety of polder landscapes of the Netherlands(from left to right): Noordoostpolder , Tzummerpolder, Polder Biesbosch , Kockengen (Paul Paris)

每一處圩田都可以看作具備自己特征的獨特景觀單元，如同一個個小塊的鑲嵌物，一齊拼成了整個荷蘭圩田景觀馬賽克。深刻理解圩田單元的獨特形式，可以讓我們了解其模式、形態學及視覺特型的復雜性——由此可以總結荷蘭低地的特征。在進行形態分析時，選取了17個代表性的圩田單元進行深入分析，以識別其圩田語法或空間句法。圩田語法指的是一套決定其空間組成的結構規律和原則。理解圩田語法是引導相關景觀轉型，或是疊加新設計于其上的起點。所選取的圩田能夠代表多樣的圩田類型，并覆蓋了荷蘭低地的不同自然地理區域，同樣也涉及到不同的圍墾時期（圖14）。

5.1分析框架

15 識別圩田語法的分析框架（S. Nijhuis提供，代爾夫特理工大學）Analytical framework for the identification of the polder grammar (S. Nijhuis, TU Delft)

16 北荷蘭省的貝姆斯特爾湖床圩田（Paul Paris攝）The Beemster lakebedpolder in North Holland (Paul Paris)

17 約1300年的自然景觀（S. Nijhuis提供，代爾夫特理工大學）The natural landscape around 1300 (S. Nijhuis, TU Delft)

在介紹每一個圩田單元時，首先描述其自然環境和歷史背景，之后對其進行詳細分析并說明其圩田語法。為了識別圩田語法，需要分析圩田形式的結構特征和表現特征（圖15）。結構特征指的是在自然景觀向可居留可開發的農業景觀轉型過程中形式上表現的特征。表現特征指的是圩田隱含的及表現出來的視覺和形式要素，這些要素促使圩田開墾過程中應用組合的、美學的和文化的主題。由此我們可以將圩田單元理解成為分層的實體，定義這一實體的則是最底層的自然景觀、進一步進行的水利干預、以及有組織的土地農業開發。三者都在形式上烙下了印記。對于這種復雜分層的識別和描述關乎問題處理，因而十分重要，可以用于解決可能存在的技術和景觀規劃上的問題，諸如在水管理和城市化方面的問題。因此，也可以獲取如何進一步開發圩田的線索。

在解讀與結構特征相關的形式系統時，對每一塊所選圩田的下述幾方面進行了系統分析并加以視覺化表達[24]。

（1）自然形式特征：自然地理位置；土壤格局、地勢、自然水系形式以及自然邊界或輪廓；（2）水利形式特征：堤壩形式、排水溝渠、排水運河及大型水道的模式、蓄水渠系統、泵水系統以及水管理單元；（3）農耕形式特征：地塊分布規律以及開發整地過程、地塊分割形式以及路網模式；（4）聚落形式特征：農舍-農莊-地塊之間的關系以及聚落的形式和分布。

研究與表現相關的特征時則分析了下述幾個方面：（1） 組織模式、規劃類型或概念；（2）風景園林化的形式：基礎的、空間的、象征性的和程序化的形式。

在發現和描述圩田的景觀特征過程中，歷史地圖、新繪制的地圖及照片是重要工具。在介紹每一處圩田單元時，首先描述其自然環境和歷史背景，之后進一步繪制基于調研的詳細地圖，包括圩田的自然景觀、地形高度、水系統、基本形式和空間形式等。

5.2以貝姆斯特圩田( Beemster)為例

18 高程示意圖（S. Nijhuis提供，代爾夫特理工大學）Height map (S. Nijhuis, TU Delft)

19 約1900年時貝姆斯特爾的水管理情況（S. Nijhuis提供，代爾夫特理工大學）The water management of the Beemster, around 1900 (S. Nijhuis, TU Delft)

下面將以貝姆斯特爾圩田作為例，對于這一框架的應用進行說明（圖16）[25]。范圍約72km2的貝姆斯特爾湖，是17世紀阿姆斯特丹北部陸地開墾活動中圍墾的一系列湖泊中面積最大的[26]。公元1608至1612年，阿姆斯特丹富商對貝姆斯特爾湖床圩田投資進行開墾。因為眾多原因，人們對其贊不絕口：作為治水成就，它宣告了荷蘭人征服自然的能力，它是精妙技術和強大組織的產物，是農耕富足的良田，是浸潤著建筑和園藝之美的知識寶庫[27]。在這處湖床圩田中，人與景觀的關系展現在其構筑系統的理性的尺寸和比例中。然而其組織形式并非單純從功能和可持續角度出發，也有著美學的吸引力。這一圩田在1999年被聯合國教科文組織收入了世界遺產名錄。

從自然形式特征來看，貝姆斯特爾圩田的輪廓直接源于一個自然泥炭湖泊，這個泥炭湖泊由先前的泥炭景觀排水后形成的沼澤河流侵蝕而成（圖17）。地表高程約為海平面海拔（ASL）之下1.9m至4m不等（圖18）。風車、環狀水渠以及圍合堤壩構成其水利形式重要特征。最初圩田中架設43座風車，分3步將水泵入環狀水渠中（圖19）?，F如今少數幾個電泵就能替代它們。圩田細分為4個水管理單元，每個單元的海拔不同，并與湖底的自然地勢相融合。四周則圍有約42km的環狀水渠（蓄水渠系統的一部分）以及堤壩。圩田地塊分割的模數系統決定了其農耕形式特征。土地測量員Lucas Jansz Sinck （未知- 1622）完成了它的設計布局（圖20）。經過幾次修改后，最終版的格網由5-6條道路和4-5條水渠構成，交點間距離約為1 880m?！皹藴实貕K”是地塊分割的最小單元，尺寸約為185×940m（圖21）。地塊的上端常常連接道路，終端則靠近排水渠。5個標準地塊合并，形成了約940×940m的方形結構，方形的兩邊是排水渠，另外兩邊則是道路。4個這樣的方形合并構成“貝姆斯特正方形”，四周是林蔭路與十字交叉的溝渠。路網模式的規律性在當時有一定的實用性，采用盡可能多地修路的方式來延續交通，因而路網保持通暢。土地主人被授予“種植權利”，因此他們自行出資，沿著內部堤肩及自有土地前道路種植成排樹木，形成了現有的典型封閉式景觀空間（圖22-23）。在聚落形式方面，北荷蘭的施托爾普農舍（stolp）是貝姆斯特爾墾田農舍的典型，農舍附有農場，四周是樹林帶。貝姆斯特爾的一些施托爾普式農舍被稱為“紳士農場”。尹荷恩農舍（the Eenhoorn）（1682）是現存的一處實例（圖24）。這一農舍因其代表性的對稱式正立面而聞名。建筑師Philips Vingboons（1607-1678）設計了這一高大的正立面，有一極具表現力的中部，在其上冠以頸狀山墻。另有約50處鄉村莊園，現多已不存。雖然對建筑的排列方式有相關規定，但可以對格網系統進行靈活多變的調整，因此網格也具有層次[28]。林蔭路、空間、村莊、規劃廣場、鄉村莊園、農場、水道以及風車結合為整體，形成了貝姆斯特爾的景觀新類型。這種類型影射出了荷蘭景觀及城市以理性方式彼此聯系的意向。圩田是自然、健康和生活幸福的荷蘭理念的象征，也代表了新荷蘭共和國的和平和富饒。同時圩田也因其擁有的田野的肥沃、鄉村莊園的舒適及花園設計的優美而飽受贊譽[29]。模塊式的方形和諧地融入了圩田之中，它們的反復出現，創造出區域無限大的幻覺。自然形式就其本身而言并不占據主導。例如其邊界，邊界僅僅存在于林蔭路系統的地平線中，因而并不明顯。因此荷蘭低地的三維性、無限平面的錯覺以及規律性的完美分割，是圩田最為重要的特性，也最容易遭到破壞。

20 貝姆斯特爾的地塊分割地圖，由Lucas Jansz Sinck于1612年繪制（Westfries檔案館提供）Parcellation map for the Beemster by Lucas Jansz Sinck, 1612 (Westfries Archive)

21 貝姆斯特爾的基本形式（S. Nijhuis提供，代爾夫特理工大學）The basic form of the Beemster (S. Nijhuis, TU Delft)

貝姆斯特爾圩田的實例證明，我們可以運用這種分析框架來研究圩田形式的結構特征和表現特征，也就是圩田語法。在每一處圩田之上，隨著其圩田語法的應用與經營，呈現出特殊的形式。圩田具備獨特個性和景觀特點，其場地精神也得以展現。在這些形式特征的基礎之上，可以從空間和時間方面來對不同的圩田單元進行比較?；谯滋锏牡乩砦恢靡约伴_墾方式，可以識別出其空間上的類似之處，時間方面則是通過開墾時期來比較。通過這種比較可以總結出一系列不同的圩田類型。其中一些是基于其自然、水利、聚落和農業形式，另一些則是基于其風景園林學特性。這些類型揭示出，在相關技術及設計手法的運用下，逐漸形成的圩田景觀的共同特性及其發展歷程。

圩田語法不僅僅是從風景園林層面理解圩田景觀的工具，同樣也是未來發展的基石。通過開展基于圩田語法的設計實驗，我們可以探討圩田景觀的哪些要素可以轉型以及其如何轉型，同樣也可以研究采用什么樣的形式可以與經濟和氣候變化相適應。風景園林設計實驗——通過設計的研究（research by design）——有助于探尋空間發展的可能性，誘導針對現有設計難題的解決措施及方案的生成[30]。這一類基于設計的知識經驗可以導向新型圩田景觀的形成，這種景觀均衡連貫，并具有自己獨特的特性和空間品質。

22 空間形式（S. Nijhuis提供，代爾夫特理工大學）The spatial form (S. Nijhuis, TU Delft)

23 貝姆斯特爾的一處典型的林蔭道（S. Nijhuis提供，代爾夫特理工大學）A typical alee in the Beemster (S. Nijhuis, TU Delft)

24 尹荷恩農舍（the Eenhoorn）（1682）因其代表性的對稱式正立面而聞名，帶有高大、極具表現力的中部，在其上冠以頸狀山墻（S. Nijhuis提供，代爾夫特理工大學）The Eenhoorn (1682) is renowned for its representative, symmetrical front fa?ade with its tall, expressive central section, crowned with a neck-gable (S. Nijhuis, TU Delft)

6　結論

圩田景觀是變化中的景觀。有些變化影響深刻，有時則不那么劇烈。應該將圩田景觀理解為一個動態的系統，這個系統受到自然過程和社會需求的影響而不斷發生變化。自然景觀由地質過程形成，并沒有受到人類的干擾。在開墾、定居和對運用技術改造自然景觀的過程中，文化景觀漸漸形成。因此，圩田是幾種條件相互作用的結果，其中包括自然景觀條件、開墾水管理技術經驗的可能性以及社會文化背景。同樣，圩田景觀是一種分層實體，時間的痕跡層層覆蓋。其相似點不斷累積，差異也不斷擴大，因而形成了多種多樣的圩田景觀，其中每一種都擁有獨特個性。然而，圩田景觀依舊很容易受到影響。由于城市飛速發展以及自身功能的轉變等，這些鄉村景觀通常遭受地平統一傾向及地塊標準化的不利影響，因而喪失了其文化特性。因此，我們有必要從風景園林的視角解讀這些景觀，以從中提取隱含的相關知識和設計經驗，并以正確的方式加以運用。

本文中呈現出了一種理解圩田景觀一致性和差異性的特別方式，通過系統化描述和空間設計原則識別，為空間規劃設計提供線索。在荷蘭，完成了首份綜合性荷蘭圩田地圖的繪制。這張地圖同樣也劃定了一項“基準”，這一基準不僅可以指導圩田景觀的歷史重建，也拓寬了當下眼界和未來視角。此外，進一步研究將圩田作為低地的基礎景觀單元研究。對選定圩田案例進行類型學分析，并分別列舉和解讀了不同圩田類型（海洋黏土圩田、河流圩田、湖床圩田以及泥炭土圩田）的特性。盡管荷蘭圩田是研究的核心，但這一研究原則在全球背景下均適用。本研究方法論體現了風景園林研究的典型思考方式，這種方式具備以下三個特征：首先，以多尺度的研究手法，將全部的圩田土地囊括在內，同時也通過圩田語法識別的方式，重點研究單個圩田。第二，研究將空間可視結構與水管理模式結合起來。第三，以有形的形-態為媒介，將圩田看作自然、生物和社會文化作用力共同作用的結果來研究。截至現在，研究人員僅僅是從水管理和歷史角度出發對圩田景觀進行探討，而這樣的分析可以使我們理解圩田的空間設計原則，并獲得圩田保護及轉型的啟示。

1　INTRODUCTION

Polders can be found in coastal and alluvial lowlands all over the world. In these reclaimed areas water levels are artificially controlled so people can live and work there. This often centuries-old interaction between man and water has produced a rich variety of polder landscapes. These landscapes are under threat due to climate and economic change. Increasing flood risk due to sea level rise, ongoing subsidence due to intense drainage, and rapid urbanization ask for action to guide these developments. These valuable cultural heritage landscapes must be safeguarded through knowledge development providing clues for preservation and transformation. Though there are many studies on the polder as hydraulic or historic phenomenon, there are hardly any who address polders as spatial constructions and cultural expressions: the polder landscape as one can see and experience it, and also as a display of cultural aspects.1To fill this gap the Chair group Landscape Architecture at Delft University of Technology (Netherlands) initiated a long term research program with the focus on the landscape architecture of polder landscapes. This program aims to develop knowledge and raise awareness of these cultural flood-prone lowland landscapes and their problematic situation of intensive cultivation and habitation, while providing clues for policy making, planning and design from a landscape architecture perspective. The research addresses polder landscapes as cultural expressions-rather than only as results of water engineering-employing mapping, case studies and comparative research as means to understand the similarities and differences of these 'polderscapes'. Key results include Sea of Land (2007), with the emphasis on lakebed polders, The Polder Atlas of the Netherlands (2009), providing a more inclusive perspective on polders, and the Comprehensive Polder Map of the Netherlands (2013), the first systematic overview of polders in the Netherlands.2Polder landscapes of the world is a recent initiative that researches polder landscapes from an international perspective.

This article aims to provide a state of the art overview of the landscape architecture research on the Dutch polder landscape. It elaborates on the methodical backgrounds and offers insight in the systematic exploration and delineation of the wealth in polderscapes throughout the scales that resulted in a GIS-based country covering map and the characterization of polder types based on representative case studies. The article starts with introducing the Dutch polder landscape as a cultural product. Next, the research strategy is introduced elaborating on morphological research and maps as tools for visual thinking and communication. Consequently, the construction of the polder map of the Netherlands will be discussed addressing the main sources and the identification of polder units. The next section addresses the analytical framework for exploring the characteristics of individual polders, which is illustrated by an application on the Beemster, an example of a seventeenth century lakebed polder in the Netherlands. The article closes with some concluding remarks.

2　THE DUTCH POLDER LANDSCAPE AS CULTURAL PRODUCT

A polder landscape can be defined as: “A level area, in its origin subject to high water levels (permanently or seasonally, originating from either ground water or surface water), but that through impoldering is separated from its hydrological regime in such a way that a certain level of independent control of its water table can be realized.3As such the resulting polder landscape is a cultural-historical expression of the local interaction between physical conditions, knowledge and management." This definition includes a wide range of polder landscapes in areas where waterlogging or inundation occurs permanently, such as: fens and bogs, shallow sea and lake beds, or temporarily, such as tidal lowlands and river plains. Based on this distinction three groups of polder landscapes can be identified: (1) impoldered low-lying lands, (2) lands reclaimed from the sea, and (3) drained lakes.4Though there is a certain similarity in these polder landscapes, there is a great variety in polder forms caused by differences in the geological subsoil, the dynamics of water and land and human intervention across the world. Though physical conditions are an important factor for these landscapes, also humans play a decisive role in their development since landscapes are the result of the action and interaction of human and natural factors. As the geographer Carl Sauer stated: “Culture is the agent, the natural area is the medium, the cultural landscape is the result."5

2.1Surveying, reclaiming and dividing territories

There is a long tradition of designing, building and maintaining productive polder landscapes in the Rhine-Meuse-Scheldt delta located in the Netherlands (Fig. 1). Polder landscapes are inherently part of the Dutch culture and have along history. In physical terms, since almost half of the country consists of polders, but also in terms of knowledge production and water management. Here the oldest water infrastructures such as dikes, dams and culverts date from pre-Roman times, but it was not until after the arrival of the Romans in the first century AD that large scale reclamations of peat fens and bogs took place.6Later, from the tenth century onwards the rest of the Dutch peat lands were systematically reclaimed. In the sixteenth century technological advances in engineering, surveying and cartography gave a new impetus to land reclamation. Not only advances in dike building (ca. 1576) and the application of windmills for pumping water (from 1533 onwards) made it possible to reclaim land from lakes. But also the invention of triangulation in the 1530s and the use of the measuring chain and compass made it possible for land surveyors to survey larger areas without too much distortion.7Surveyors also started to make maps for inventorying, land division (design of layouts), land management or improvement of land and therefore played an crucial role in dike building, hydraulic works and large land reclamation projects. In that time period maps also became legal tools as land became a commodity with the emergence of capitalism. Maps as representation helped to create an image of community built on civic virtue and territorial reclamation.8

The potential for building and farming made lakebed polders economically desirable. Wealthy merchants invested in these vast projects for the creation of fertile land to provide food for the cities and to create pleasant areas to live avoiding the pollution of the cities. Lakes had to be transformed and cultivated into a practical economic environment, with beautiful estates and profitable husbandry. By 1640, Amsterdam had forty percent more land than in the sixteenth century land due to lakebed polders such as the Beemster (1608-1612), Purmer (1622), Wormer (1625) and Schermer (1631). The layout of these polders was determined by surveyors that were trained in practical methods for land surveying, mapping and drawing. The polder layout was not the result of a single design, but emerged from a practical division of land in multi-layered grids of suitable rectangular blocks that adapted to the local soil composition, hydraulic engineering, optimal land use, etc. These surveyors built on centuries of experience with the application of various and flexible types of grids that could be adapted easily to local circumstances.9Simon Stevin (1548-1620), who stood at the basis of Dutch theories of urban planning, settlements and fortifications, adapted these principles and codified flexible and temporal planning applications of these multi-layered grids with symmetry, axial organization, and modular systems of rectangles for urban development.10

2.2Garden of Holland

Land reclamation, cultivation and the creation of this typical geometrical landscape lay at the foundation of the art of gardening in Holland, so much that the country itself became identified with the garden and its people with gardeners: the Garden of Holland or Hortus Batavus (Fig. 2).11The rectilinear patterns created by the seventeenth century reclamation techniques and land surveying turned reclaimed land into a truly cultural landscape in which country estates were built for profit and ornament.12The geometric order of their accompanying gardens fitted perfectly into the regular lines of the polder landscape.13This may be exemplified by the royal garden of Frederik Hendrik (1584-1647) at Honselaarsdijk. This garden became a representation of the Garden of Holland with the House of Orange as its protector (Fig. 3).14Honselaarsdijk was, as many other estates, integrated in the surrounding polder landscape, carefully adjusting to specific topographic conditions and employing common methods of land division and reclamation. The result was a combination of the practical and the pleasurable and symbolized the improvement of nature by man: the garden was a polder and became part of the rhetoric surrounding of the Garden of Holland.

In the Dutch situation the sixteenth and seventeenth were decisive in the cultural formation. Impoldering became a form of cultural production where politics, economics, and cartography came together. The regularized layout of the polder landscape paralleled contemporary ideas about civilization, ownership and statecraft. Though the term "polder" originates from the Netherlands (17th century: "polre"), that does not mean that polder landscapes as such are invented by the Dutch. From the dawn of human kind people are already taking advantage of land-water conditions to create productive landscapes while constructing water infrastructures for flood protection (dikes and dams) and water management (drainage or irrigation) in Mesopotamia from the third millennium BC onwards.15However, in the sixteenth and seventeenth century the Dutch cultivated the idea of polder-making. The polderbecame the nation’s symbol for taming nature in the struggle with the elements of water and land with a strong will and technological expertise. At the same time became land reclamation also a Dutch export product exemplified by their involvement in poldermaking across Europe and the rest of the world from the 12th century onwards until now.16

The Dutch polder landscape of today is gradually transformed into multifunctional spaces where uses such as leisure and tourism, nature, water storage and housing become more and more important besides agriculture. These developments put pressure on the quality of the space and the identity of the rural landscape. The key to solving many spatial issues lies in the wealth of shapes of the polder landscape itself.

3　UNDERSTANDING THEPOLDER LANDSCAPE

Morphological research is employed to study the Dutch polder landscape. Through exploring the morphology of the polder landscape the genius loci of the Dutch lowlands can be 'read'in order to retrieve the information and design knowledge that lies hidden within it, as clues for further development. Morphology or formal studies have proven to be useful in such a way that the landscape can be read and analysed via the medium of its physical form. The focus is on the tangible results of physical, biological and sociocultural forces. Formal analysis is an interpretative research method (also called: hermeneutics), in which knowledge is acquired through formal analysis and interpretation of visual representations of landscapes. The form is regarded a purely evidential system without prepossession regarding the meaning of its evidence, and presupposes a minimum of assumption.17This type of research is not only directed towards gaining knowledge on the architecture of the landscape, but is also focussed on cultivating spatial intelligence, the capacity to understand and design spatial compositions and relationships across scales. Morphology studies have three characteristics.18Firstly, they consider all scales of landscape, from gardens to regional landscapes. Secondly, they combine volumetric characteristics of landscape structures with patterns of space and movement. Thirdly, they are a morphogenetic approach, since the landscape architectonic composition is a product of time-the time of its conception, development or mutation. Morphology studies can also be found in urban design, architecture, art and geography.19

In landscape morphological research maps are considered to be fundamental tools for knowledge discovery. Maps have served as vehicles to represent and understand the landscape for centuries. Maps facilitate a spatial understanding of things, concepts, conditions, processes or events in the human/natural world.20They are informative as well as persuasive and are therefore important means of knowledge acquisition. This can be distinguished from knowledge acquisition through processing of spatial information that is termed mapping. Mapping refers to a process rather than a completed product. Mapping entails cartographic exploration, an activity that exploits the agency of maps and map-making to construct and communicate spatial knowledge in a visual way. In this respect mappings are not one-to-one copies of the existing situation, but their production require interpretation and reformulation. Mapping allows to 'digest' information in a rational and systematic way, which is a personal process influenced by the choices and judgments made by the interpreter, referred to as visual thinking. At the same time, these findings are made transferable via visual representation, which showcases relationships, structures and patterns, referred to as visual communication. Map dissection, map comparison, and map addition analysis are useful analytical operations.21These cartographic explorations help researchers in landscape architecture to acquire new or latent information, which is the basis for generating spatial knowledge that informs the design process. It facilitates interpretation of data to become knowledge, which serves as a base for invention through design (Fig. 4).

The morphology of the polder landscape is analysed in two stages, revealing its cohesion and variety through systematic analysis and cartography. First, all polders are surveyed and digitized resulting in the first comprehensive polder map of the Netherlands. Second, the polder as the elementary landscape unit of the lowlands is further explored.

4　COMPREHENSIVE POLDER MAP OF THE NETHERLANDS

The polder map of the Netherlands systematically visualizes the wealth in shapes of the polder landscape (Fig. 5). It is unique in a sense that it embraces the whole of the Dutch polder landscape focusing on the polder as a hydraulic and landscape entity. Though the Netherlands has a long tradition of polder cartography, with unique polder maps from individual polders or polder districts maintained by waterboards, national polder maps are lacking (Fig. 6). The polder landscape ofthe Netherlands is often described in books and mainly from the water management or historical point of view. When A.A. Beekman's Netherlands as polder land (in Dutch) appeared in 1884, he laid the first scientific foundations for the description of the Dutch polder land. Others followed in his footsteps, focusing on specific regions or lakebed polders in particular. The comprehensive polder map of the Netherlands is the first attempt on the national scale that considers the polder as a hydraulic construction and as a structure in the landscape. Insight into this connection is essential for an understanding of the polder as a spatial form with its own signature. Therefore the polder map, as a nation-wide map in which this connection is made readable, is a new step in Dutch polder cartography.22

Since the Dutch landscape changed dramatically since the beginning of the twentieth century, due to explosive urbanization and economic development, it is hardly possible to use contemporary topographic maps to identify polder units, particularly the older ones. To find a solution for this the practically intact state of the landscape, combined with the availability of detailed map material, were the most important considerations. Historical maps from the period between 1800 and 1900 offered the best opportunities for assembling a complete polder map. In this period the cities -with several exceptions-were still within their old fortifications. The landscape, as it had come to exist in the preceding centuries, could still be read in a largely unaltered state. In the 1800s surveying had already attained great precision, so that map images from the time could be projected onto contemporary maps without many problems. This was also the period when a start was made with the production of a country covering Topographic and Military Map 1850-1864 (TMK) and the Hydraulic Map first edition 1865-1891, both with a scale of 1:50.000 (Fig. 7 and 8). These two analogue historical maps served as the basis for the construction of the polder map. Further the highly accurate modern Height map of the Netherlands (AHN5), an airborne LiDAR point cloud (8 points per m2), was employed to ensure precise delineation and to identify recent polders (polders constructed after 1850) (Fig. 9).

The first step in the construction of the polder map consisted of digitalization and georeferencing of the in total 366 map sheets of the Topographic and Military Map 1850-1864 and Hydraulic Map first edition 1865-1891. The maps were scanned and transformed into a digital file consisting of pixels. The digital maps sheets were geo-rectified and georeferenced by means of a Geographic Information System (GIS). In this way the historic maps were corrected in planimetric terms and augmented with geographic coordinates related to the Dutch standard coordinate system.23This enables precise combination and exchange of geo-referenced information.

The next step was to interpret the maps, to identify the polders as spatio-hydraulic landscape units and to vectorise them (Fig. 10). The Height map of the Netherlands was used to evaluate and correct the drawing and also to identify land reclamations after 1850. The result is a GIS-based polder map of the Netherlands with 9.087 identified polder units represented on a scale of about 1:10.000 to 1:25.000, and includes all types of polders: impoldered low-lying lands, lands reclaimed from the sea, and drained lakes. More specific, the map shows all hydraulic units together with their specific boundaries. Within this framework, the polders are designated as spatial units.

The use of GIS not only ensured precision work, but also makes it possible to link information to the map, turning it into a spatial database that includes: names, year of impoldering, numerical data, etc. Based on the Geographical Dictionary of the Netherlands (Van der Aa, 1839-1851, in Dutch), a database was created with all the polders that are mentioned in this dictionary (Fig. 11). This information made it possible to supplement and check the data from the TMK and Hydraulic Map. By applying GIS the information can become spatial and available for analysis and visualization. It enabled for instance to study the development of the polder landscape in time (Fig. 12), but also computational analysis of polder properties such as: polder size, form complexity and polder form.

In order to get a grip on the wealth of polder forms the polder map was interpreted based on physical-geographical characteristics derived on differences in the substrata (rock and relief), soil and hydrology. Anthropogenic influences also play a role in this, so that, as a result of this interplay between the abiotic and anthropogenic, particular combinations of characteristics can be typical. The result is a polder typology based on physical-geographical landscapes, each of which has a number of salient formal characteristics (Fig. 13). Polder types include: sea clay polders (e.g. Tzummerpolder), peat polders (e.g. Ronde Hoep), river polders (e.g. Alblasserwaard) and IJsselmeerpolders (e.g. Noordoostpolder). Lakebedpolders (e.g. Beemster) are regarded a subtype of peat polders since they originate from peat lakes.

5　IDENTIFICATION OF POLDER GRAMMAR

Each polder can be regarded as a unique landscape unit with its own characteristics, one tesserae which goes to make up the mosaic of the whole Dutch polder landscape. Insight into the peculiar form of the polder unit provides insight into the complexity of the pattern, the morphology, the visual qualities-and with that, the identity of the Dutch lowlands. In a morphological analysis seventeen selected representative polder units are further analyzed in order to identify their polder grammar or spatial syntaxes. The polder grammar is the set of structural rules and principles that determine the spatial composition. Knowledge of the polder grammar is the starting points for new transformations of the landscape involved, or adding a new design layer. The selected polders represent the wide variety of polder types across the different physical-geographic regions in the Dutch lowlands, as well as the different time periods of impoldering (Fig. 14).

5　.1 Analytical framework

Each polder unit is introduced through description of the physical setting and the historical backgrounds followed by a detailed analysis and description of its polder grammar. In order to identify the polder grammar the constructive characteristics and the expressive characteristics of the polder form are analysed (Fig. 15). The characteristics of construction refer to the formal aspects of the transformation of the natural landscape into a habitable and exploitable agricultural landscape. The characteristics of expression refer to the implicit and explicit visual and formal elements that bring compositional, aesthetic and cultural motifs of polder making into play. In this way the polder unit is understood as a layered entity defined by the underlying natural landscape, and to a further degree by the hydraulic interventions, and finally by the organization imposed in the agricultural development of the land. Each stage leaves its traces on the form. The identification and delineation of this complex layeredness is important because this creates a point of contact for dealing with technical and landscape planning questions which may be present, such as water management or urbanization, and thus provides clues for further development of the polder landscape.

To unravel the formal systems related to the construction the following aspects are systematically analyzed and visualized for each selected polder:24

(1)Natural formal characteristics: physicalgeographic location; soil pattern, relief, natural drainage pattern and natural boundaries or contours;

(2)Hydraulic formal characteristics: form of the dikes, the pattern of drainage ditches, drainage canals and larger watercourses, the boezem system, the pumping system and water management units;

(3)Agricultural formal characteristics: the distribution of plots and their preparation for exploitation, the form of parcellation and the pattern of the roads;

(4)Formal characteristics of settlement: the farmhouse-farmstead-parcel relationship and the form and distribution of the settlements.

Regarding the characteristics related to the expression the following aspects are analyzed:

(1)Organizational model, plan type or concept;

(2)Landscape architectonic form: the basic, spatial, symbolic and programmatic form.

Historical maps, newly drawn maps and photographs are employed as important means to discover and delineate their landscape architectonic characteristics. Each polder unit is introduced through description of the physical setting and the historical backgrounds followed by detailed maps that delineate the natural landscape, terrain height, water system, basic form and spatial form of the polder under investigation.

5.2The Beemster polder as an example

In order to illustrate the application of this framework the Beemster polder is used as an example (Fig. 16).25The Beemster Lake, seventytwo square kilometres in extent, was the biggest of a series of lakes north of Amsterdam drained in the seventeenth century for land-based development.26The resulting Beemster lakebed polder was constructed between 1608 AD and 1612 AD on behalf of wealthy merchants from Amsterdam. It has since been hailed for many different reasons: as a triumph over water, a statement of Dutch power over nature, a product of technical ingenuity and organizational prowess, a site of agricultural abundance and a repository of architectural and horticultural beauty.27In this lakebed polder the relation between man and the landscape was fulfilled in the dimensions and proportions of a rational, architectonic system. The organization wasnot only functional and sustainable, but at the same time aesthetically appealing. Since 1999 the polder is designated as Unesco world heritage.

Regarding the characteristics of the natural form the contour of the Beemster is directly derived from the form of a natural peat lake, which was developed by erosion of a bog river that drained the former peat landscape (Fig. 17). The ground level varies from 1.9 to mostly 4 metres below ASL (Fig. 18). Important characteristics of the hydrologic form are determined by the windmills, ring canal, and surrounding dike. Originally 43 windmills pumped the water up in three steps into the ring canal (Fig. 19). Nowadays this is done by only a few electrical pumps. The polder was subdivided into four water management units that differed from one another in level, coordinated with the natural relief of the lake bottom. The ring canal (part of the boezemsystem) and the surrounding dike have a length of about forty-two kilometres. Agricultural formal characteristics are determined by a modular system of parcellation and it was the land surveyor Lucas Jansz Sinck (unknown-1622) who designed its layout (Fig. 20). After several changes the final grid consisted of five to six roads and four by five ditches intersected at distances of about 1.880 metres. The smallest unit in the parcellation is the 'standard parcel' of about 185 x 940 metres (Fig. 21). The head of the parcel is always connected to an access road, and the foot with a drainage ditch. Five standard parcels together form a square of about 940 x 940 metres, with drainage ditches on two sides and roads on the other two. Four of these squares form the 'Beemster square', enclosed by roads with trees and intersected by a cross of canals. The regularity of the road pattern did have a certain utility in that time, by spreading traffic over as many roads as possible, the roads remained passable. The owners were granted 'planting rights', whereby at their own expense they planted trees along the inside shoulder of the dike and along the roads in front of their parcels, nowadays determining this typical enclosed landscape spaces (Fig. 22 and 23). Regarding the settlement form the North Holland stolp, with a farmyard surrounded by a belt of trees, is the typical reclamation farmhouse of the Beemster. Some of the stolpen in the Beemster were 'gentlemen's farms'. An example, the Eenhoorn (1682) still exists (Fig. 24). The farmhouse is renowned for its representative, symmetrical front fa?ade with its tall, expressive central section, crowned with a neck-gable, a design by the architect Philips Vingboons (1607-1678). Also there were about fifty country estates, of which most are demolished. There were regulations about the alignment of the building lots but the grid system allowed for some flexibility and variety resulting in a hierarchy of grids.28

The avenues, rooms, villages, projected squares, country estates, farmhouses, watercourses, and windmills of the Beemster together formed a new type of architectonic landscape in which the image of the Dutch landscape and the Dutch city were linked to one another in a rational arrangement. The polder came to epitomise Dutch ideas of pristine nature, wholesome and blissful living, just as it symbolised the peace and wealth of the new Dutch Republic. It was celebrated for its pastoral richness, its pleasant country estates and beautifully designed gardens.29The modular square is harmoniously embedded in the polder, and by its repetition creates the illusion of an infinite area. The experience of the natural form in itself, represented by the edge, plays no significant role. The edge is only the horizon for the system of avenues. This three-dimensionality, the illusion of an infinite plane, divided with perfect regularity that suggests the Dutch lowlands, is the most important but also the most vulnerable quality of this polder.

As exemplified by the Beemster polder the analytical framework can be used to study the constructive and expressive characteristics of the polder form, its polder grammar. In each polder the application and elaboration of the formal grammar laid out above has led to a specific form and a unique character and landscape identity in which the genius loci of the polder is captured. On the basis of these formal characteristics the various polder units can be compared with one another both spatially and temporally. Spatial similarities can be identified on the basis of geographic location and the way in which the polders were created, and temporal similarities on the basis of the period in which they were created. Various typological series can be derived from this comparison of the polders. Typological series based on the natural, hydraulic, settlement and agricultural form and also typological series based on landscape architectonic features. These series reveal the common characteristics and lines of development along which the polder landscape evolved, in terms of both the technology involved and the design.

The polder grammar is not only a vehicle for understanding the landscape architecture of the polder landscape, it can also act as the basis for future development. Through design experiments based on the polder grammar one can investigatewhich and how elements from the polder landscapes can be transformed, and in what form to adapt to economic and climate change. Landscape architecture design experiments - research by design - can help to explore the possibilities for spatial development, generating proposals or potential solutions for design problems.30This type of knowledge based design leads to new, balanced and coherent polder landscapes with their own identity and spatial qualities.

6　IN CONCLUSION

Polder landscapes change. Sometimes there are profound changes taking place, sometimes less radical changes. Therefore the polder landscape should be understood as a dynamic system that continually transforms under the influence of natural processes and social requirements. The natural landscape was formed by geological processes without human intervention. The cultural landscape emerged from the reclamation, colonization and technical mastery of the natural landscape. As such is the polder a result of the interaction between the conditions offered by the natural landscape, the technical possibilities for reclamation and water management, and the sociocultural context. As such is the polder landscape a layered entity where traces that time has laid over can reinforce or contradict each other with a wide variety of polderscapes as a result, each with their own identity. However, the polder landscape remains subject to change. Due to rapid urban development, function change, etc. these rural landscapes often suffer from levelling tendencies and standardization that negate the characteristic spatial differences with the loss of cultural identity as a result. Therefore it is necessary to understand these landscapes from a landscape architecture point of view in order to retrieve the information and design knowledge that lies hidden within it and then apply them in the right way.

This article presented a particular way of understanding the coherence and variation of the polder landscape providing clues for spatial planning and design through systematic delineation and the identification of spatial design principles. As exemplified by the Dutch case this resulted in the first comprehensive polder map of the Netherlands. The map provides both an overview and a ‘benchmark’, a point of reference from which the history of the polder landscape can be reconstructed and that provides a perspective of the present and the future. Also the polder as the elementary landscape unit of the lowlands is further explored. In a morphological analysis of selected examples the characteristics of the different polder types (sea clay polders, river polders, lake bed polders and peat polders) are enumerated and interpreted typologically. Though the Dutch polder landscape played a central role here, the principles of study are applicable in a global context. The presented methodology embodies a way of thinking typical for landscape architecture and has three characteristics: First, it entails a multi-scalar approach covering the polder land in total but also addresses individual polders through the identification of polder grammar. Second, it combines spatio-visual structures with patterns of water management. Third, it explores the landscape via the medium of its tangible form as a result of physical, biological and socio-cultural forces. An analysis of this sort makes the polder landscape, which until now often is discussed only in water management and historical terms, accessible for spatial design disciplines while providing clues for preservation and transformation.

1 Examples from the water management and historical perspective include: Rippon, S. (2000) The transformation of coastal wetlands. Exploitation and management of marshland landscapes in North West Europe during the Roman and Medieval periods. New York: Oxford University Press; Van der Ven, (2004) Man-made lowlands. History of water management and land reclamation in the Netherlands. Utrecht: International Commission on Irrigation and Drainage. An example from the cultural perspective: Wagret, P. (1968) Polderlands. London; Methuen & Co.

2 Reh, W., Steenbergen, C.M. & Aten, D. (2007) Sea of Land. The polder as an experimental atlas of Dutch landscape architecture. Wormer: Uitgeverij Noord-Holland; Steenbergen, C. M., Reh, W., Nijhuis, S., & Pouderoijen, M. T. (2009) The Polder Atlas of the Netherlands. Pantheon of theLowlands. Bussum: THOTH publishers; Nijhuis, S & Pouderoijen, MT (2013) De polderkaart van Nederland. Een instrument voor de ontwikkeling van het laagland (The poldermap of the Netherlands. An instrument for spatial development of the lowlands). Bulletin KNOB 3; 137-151. DOI: 10.7480/knob.112.2013.3.626

3 Centre for Civil Engineering Research and Codes (CUR) (1993) Hydrology and water management of deltaic areas. Rotterdam: Balkema, p. 230.

4 Cf. CUR 1993 (note 3), p. 230.

5 Sauer, C.O. (1963) ‘The morphology of landscape’. In J. Leighly (Eds.) Land and life. A selection from the writings of Carl Ortwin Sauer(pp. 315-350). Berkeley, CA: University of California Press. (Original article published 1925), p. 327.

6 See for an overview of the oldest known water infrastructures in the Netherlands: Lascaris, M. & De Kraker,

A. (2013) ‘Dikes and other hydraulic engineering works from the Late Iron Age and Roman Period on the coastal area between Dunkirk and the Danish Bight’. In: Thoen, E. et al. (eds.) Landscapes or seascapes? The history of the coastal environment in the North Sea Area reconsidered (pp. 177-198). Turnhout, Brepols.

7 See for a history of Dutch surveying and cartography: Koeman, C. & Van Egmond, M. (2007) ‘Surveying and official mapping in the Low Countries, 1500-ca. 1670’. In: Woodward, D. (Ed.) The history of cartography. Volume III: Cartography in the European Renaissance (pp. 1246-1295). Chicago: Chicago University Press.

8 Sutton, E. (2015) Capitalism and cartography in the Dutch golden age. Chicago: University of Chicago Press.

9 Van der Heuvel, C. (2011) ‘Multilayered grids and Dutch town planning. Flexibility and temporality in the design of settlements in the low countries and overseas’, In: Lombaerde, P. & Van de Heuvel, C. (eds.) Early modern urbanism and the grid (pp. 27-44). Turnhout: Brepols, p. 44.

10 Van der Heuvel 2011 (note 9), p. 44.

11 Bezemer-Sellers, V. (2001) Courtly gardens in Holland, 1600-1650. Amsterdam: Architectura& Natura Press. p. 9.

12 De Jong, E. (2000) Nature and art. Dutch garden and landscape architecture, 1650-1740. Philadelphia: University of Pennsylvania Press, p. 18.

13 De Jong 2000 (note 12), p. 21.

14 Bezemer-Sellers 2001 (note 11), p. 21.

15 Viollet, P.L. (2007)Water engineering in ancient civilizations: 5,000 years of history. Madrid: International Association of Hydraulic Engineering and Research.

16 See for the role the Dutch played in polder-making around the globe in a historical and modern perspective: Veen, J. v. (1952). Dredge, Drain, Reclaim. The art of a nation. The Hague; Segeren, W.A. et al. (1983) Polders of the world. Final report. Wageningen; Danner, H. et al. (2005) Polder Pioneers. The influence of Dutch Engineers on water management in Europe. Utrecht.

17 Sauer 1963 (note 5), p. 327.

18 Adapted from Moudon, A. V. (1994)‘Getting to know the built landscape: Typomorphology’. In K. A. Frank, & L. H. Schneekloth (Eds.), Ordering space. Types in architecture and design (pp. 289-309). New York, NY: Van Nostrand Reinhold.

19 For an overview of morphological approaches see: Nijhuis, S. (2015) GIS-based landscape design research. Stourhead landscape garden as a case study. Delft: A+BE, p. 48ff.

20 Harley, J. B., & Woodward, D. (Eds.) (1987) The history of cartography. Vol. I: Cartography in prehistoric, ancient and medieval europe and the mediterranean. Chicago and London: University of Chicago Press, p. xiv.

21 Nijhuis, S. & Pouderoijen, M.T. (2014) ‘Mapping urbanized deltas’, in: Meyer, H & Nijhuis, S (eds.) Urbanized deltas in transition (pp. 10-22). Amsterdam: Techne Press.

22 See for a full elaboration: Nijhuis & Pouderoijen 2013 (note 2) and Steenbergen et al. 2009 (note 2), pp. 74 ff.

23 See for more backgrounds on the use of historical maps, digitalization, georeferencing, etc.: Nijhuis 2015 (note 19), p. 109 ff.

24 See for backgrounds: Steenbergen et al. 2009 (note 2), pp. 28-29.

25 For a more elaborate description of the constructive and expressive characteristics of the Beemster polder and other polders see: Steenbergen et al. 2009 (note 2), pp. 338-353, pp. 198-469.

26 Van der Ven 2004 (note 1).

27 Fleischer, A. (2007) ‘The Beemster Polder: Conservative Invention and Holland'sGreatest Pleasure Garden’. In: Roberts, L. et al. (eds.) The mindful hand. Inquiry and invention from the Late Renaissance to Early Industrialization (pp. 145-166). Assen: Royal van Gorcum.

28 Van der Heuvel 2011 (note 9), p. 44.

29 Fleischer 2007 (note 26).

30 Nijhuis, S, & Bobbink, I (2012) ‘Design-related research in landscape architecture’, J. Design Research 10 (4), 239-257. DOI: 10.1504/JDR.2012.051172.

Polderscapes: The Landscape Architecture of the Dutch Lowlands

The Dutch lowlands consist mainly of polders, areas where water levels are artificially controlled so people can live and work there. This centuries-old interaction between man and water has produced a rich polder landscape. The great variety in polder forms is caused by differences in the geological subsoil, the dynamics of water and land and human intervention. This research systematically explores and visualizes the wealth in shapes of the Dutch polder landscape from a landscape architecture perspective. Polders are not just regarded as hydraulic phenomena but also as spatial constructions and cultural expressions: the polder landscape as one can see and experience it, and also as an expression of Dutch culture. Through exploring the morphology of the polder landscape the genius loci of the Dutch lowlands can be ‘read’ in order to retrieve the information and design knowledge that lies hidden within it, as clues for further development. In this research the landscape architectonic form of the polder landscape is analyzed in two stages revealing its cohesion and variety through systematic analysis and cartography. First, all polders are surveyed and digitized resulting in the first systematic polder map of the Netherlands. The use of Geographic Information Systems (GIS) not only ensured precision work, but also made it possible to link information to the map, turning it into a spatial database. Second, the polder as the elementary landscape unit of the lowlands is further explored. In a morphological analysis of selected examples the characteristics of the different polder types (sea clay polders, river polders, lake bed polders and peat polders) are enumerated and interpreted typologically. An analysis of this sort makes the polder landscape, which until now has been discussed only in physical and historical-geographical terms, accessible for spatial design disciplines while providing clues for preservation and transformation.

Polder Landscapes; Design Analysis; Dutch Lowlands; Formal Analysis; Typology; Mapping

TU986

A

1673-1530（2016）08-0038-20

10.14085/j.fjyl.2016.08.0038.20

2016-04-23

2016-07-25

斯蒂芬·奈豪斯/男/助理教授/博士 /風景園林研究項目負責人/荷蘭代爾伏特理工大學（TUD）建筑與建成環境學院/研究方向：風景園林研究、圩田設計、三角洲景觀、基于景觀的區域設計和GIS在風景園林規劃設計領域的應用

Author:

Dr. Steffen Nijhuis is Assistant Professor of Landscape Architecture and Team Leader Landscape Architecture Research Programme at the Faculty of Architecture and the Built Environment, Delft University of Technology (theNetherlands). He is an expert in landscape architecture research and design of polder and delta landscapes, landscape-based regional design approaches to urban development, and GIS-applications in landscape planning and design.