朱釗岑,邵明安,3,趙春雷,賈小旭,王 嬌
?農(nóng)業(yè)生物環(huán)境與能源工程?
不同盆栽基質(zhì)水分特征曲線的對比分析與模擬
朱釗岑1,2,邵明安1,2,3,趙春雷1,3※,賈小旭1,3,王 嬌1,3
(1. 中國科學院地理科學與資源研究所黃河三角洲現(xiàn)代農(nóng)業(yè)工程實驗室,北京 100101;2. 西北農(nóng)林科技大學資源環(huán)境學院,楊凌 712100;3. 中國科學院大學資源與環(huán)境學院,北京 100190)
現(xiàn)有關于盆栽控水模擬土壤干旱條件的試驗中多采用含水率作為水分脅迫閾值,然而由于基質(zhì)配比不同導致含水率相同的基質(zhì)的水分狀況也不盡相同,這導致各研究間結果難以對比和參考。為快速獲取盆栽基質(zhì)水分特征曲線,建立基質(zhì)水分特征曲線預測模型。該研究以盆栽控水試驗常用的泥炭土、蛭石和珍珠巖為基質(zhì)材料,測定了不同配比基質(zhì)的水分特征曲線,通過不同方法(多元回歸模型、人工神經(jīng)網(wǎng)絡)建立了其預測模型。結果表明,人工神經(jīng)網(wǎng)絡模型對泥炭土-蛭石復配基質(zhì)水分特征曲線的預測精度高于多元回歸;相較于人工神經(jīng)網(wǎng)絡,多元回歸模型的穩(wěn)定性更高。綜合考慮模型的精度和穩(wěn)定性,多元回歸模型是預測作物盆栽基質(zhì)水分特征曲線的最佳模型,預測精度2≥0.950,平均誤差接近0。該模型為基質(zhì)水分特征曲線快速獲取以及相關作物干旱脅迫研究間的對比提供了方法和依據(jù)。
模型;基質(zhì);水分脅迫;盆栽試驗;水分特征曲線;傳遞函數(shù)
水分脅迫閾值是判斷作物水分脅迫響應的重要指標,盆栽人工控水模擬土壤干旱條件,是確定作物水分脅迫閾值的重要方法[1-2]?,F(xiàn)有研究中的水分脅迫閾值大多采用含水率指標,對基質(zhì)水勢測定較少。盡管部分研究對基質(zhì)水勢進行了測定,但其僅用于判斷補水時機和維持基質(zhì)水分恒定,并未將其作為水分脅迫指標[3-4]。作物對水分的吸收主要與根系和土壤中水分的能量狀態(tài)(水勢)有關[5]。在不同的盆栽基質(zhì)中,同一含水率往往對應著不同的水勢[6]。這就導致了以含水率形式表達的水分脅迫閾值在不同研究間難以統(tǒng)一,不僅不利于準確認識水分在土壤和根系間的傳輸過程,還限制了不同研究結果間的對比和參考[7-9]。水分特征曲線反映了基質(zhì)水分數(shù)量和能量之間的關系。利用水分特征曲線,將含水率轉(zhuǎn)換為水勢,是解決上述問題的有效方法。
在盆栽控水試驗中,泥炭土、蛭石和珍珠巖是當前應用最廣泛的基質(zhì)復配材料[10]。不同基質(zhì)材料具有相對固定的孔隙分布特征與持水特性,作物盆栽基質(zhì)在不同水勢下的持水量主要取決于復配基質(zhì)所用的材料及其配比[11-12]。然而,不同研究中的基質(zhì)配比往往存在差異,導致其盆栽基質(zhì)水分特征曲線不同[6]。即便少數(shù)研究進行了測定,這些水分特征曲線也僅適用于特定基質(zhì),不具有通用性[13-14]。傳遞函數(shù)模型是快速獲取水分特征曲線的有效方法。在土壤物理學領域,眾多學者利用土壤顆粒組成、有機質(zhì)含量、飽和導水率等參數(shù),成功建立了不同類型、質(zhì)地土壤的水分特征曲線預測模型[15-17]。相較于土壤,盆栽基質(zhì)的組成更為簡單[10]。若能夠建立不同組成和配比的基質(zhì)的通用水分特征曲線預測模型,則有望獲得一條快速獲取盆栽基質(zhì)水分特征曲線的有效途徑。
綜上,為了建立不同配比基質(zhì)的水分特征曲線預測模型,實現(xiàn)不同研究間水分脅迫閾值由含水率向水勢的轉(zhuǎn)換,本研究采用3種常用的基質(zhì)復配材料(泥炭土、蛭石和珍珠巖),開展不同配比下的盆栽試驗,通過測定不同配比基質(zhì)的水分特征曲線,分析基質(zhì)配比與水分特征曲線參數(shù)間的關系,最終建立水分特征曲線的預測模型,以期為基質(zhì)水分特征曲線的快速獲取及不同研究間水分脅迫閾值的轉(zhuǎn)換、對比提供方法和依據(jù)。
本研究選取泥炭土、蛭石(粒徑<2 mm)和珍珠巖(粒徑<2 mm)3種基質(zhì)材料,根據(jù)不同研究中常見的復配比例[7-9],共設置11個處理。不同處理3種基質(zhì)材料的體積比詳見表1。
表1 基質(zhì)材料配比體積比設計
將復配好的11種基質(zhì)分別填入環(huán)刀后置于水中浸泡24 h,每個處理3個重復,共計33個樣品。樣品的水分特征曲線采用利用離心機法測定。將飽和后的土樣放到離心機中,通過設置離心機(HITACHI CR21G,日本)的轉(zhuǎn)速以實現(xiàn)不同的水吸力環(huán)境,待測試樣水分達到平衡后,取出樣品并稱取其質(zhì)量。測試結束后,利用烘干法(105 ℃,12 h)測定基質(zhì)干質(zhì)量,計算不同水吸力下對應的體積含水率。
1.3.1 水分特征曲線與當量孔徑
在眾多描述水分特征曲線的模型中,van Genuchten模型(VG模型)是最為常用的水分特征曲線擬合模型,該模型具有廣泛的適用性,對于不同介質(zhì)均具有較高的擬合精度。因此,本研究采用VG模型對不同處理基質(zhì)實測的水分特征曲線進行擬合,VG模型表達式如下[18]:
式中為基質(zhì)含水率,cm3/cm3;θ為飽和含水率,cm3/cm3;θ為殘余含水率,cm3/cm3;為基質(zhì)勢,kPa;、和為經(jīng)驗參數(shù),為進氣值的倒數(shù),cm-1,和為水分特征曲線形態(tài)學參數(shù),=1?1/。
基質(zhì)的持水能力與其孔隙分布特征密切相關,當量孔徑與水吸力的關系為[19]:
式中為當量孔徑,mm;為水吸力,hPa。
根據(jù)當量孔徑的大小,將基質(zhì)當量孔徑劃分為3類:通氣孔隙(>9×10-3mm)、毛管孔隙(9×10-3~2×10-4mm)和無效孔隙(<2×10-4mm),分別對應相應的水吸力分別為<33、33~1 500、>1 500 kPa:其中,<33和>1 500 kPa基質(zhì)水吸力分別對應重力水和無效水,33~1 500 kPa基質(zhì)水吸力對應毛管水[20]。
1.3.2 水分特征曲線預測模型的建立
本研究擬在11種基質(zhì)材料水分特征曲線的基礎上,利用不同配比基質(zhì)水分特征曲線的實測數(shù)據(jù)(水勢、含水率),采用數(shù)學回歸(多元回歸模型)和機械學習(人工神經(jīng)網(wǎng)絡)的方法,建立不同配比基質(zhì)的水勢與含水率預測方程。再通過bootstrap方法將不同配比基質(zhì)的水勢與含水率數(shù)據(jù)隨機分為建模數(shù)據(jù)(80%,108組水勢與含水率數(shù)據(jù))與驗證數(shù)據(jù)(20%,28組水勢與含水率數(shù)據(jù)),分別用于模型的建立與驗證。在建模過程中共進行30次數(shù)據(jù)隨機分組,用于評價模型的預測精度和穩(wěn)定性。
多元回歸指利用線性回歸方程定量分析因變量與多個自變量間的關系[21]。在本研究中,多元回歸模型的表達式為
式中(R)為水吸力為時處理基質(zhì)的體積含水率,cm3/cm3;R為處理中泥炭土、蛭石和珍珠巖的體積百分比;、、為系數(shù),為常數(shù)項。1、2和3分別為蛭石、泥炭土和珍珠巖的水分特征曲線。
人工神經(jīng)網(wǎng)絡模型是一種模擬人腦結構和功能,由大量節(jié)點相互連接而成的大規(guī)模信息處理系統(tǒng)[22-23]。由輸入層、隱藏層與輸出層三部分組成,其中輸入層負責接受外界的信息,并將其作為模型的輸入信息傳遞給隱藏層;隱藏層是模型的內(nèi)部信息處理層,負責數(shù)據(jù)處理與轉(zhuǎn)換并將結果傳遞給輸出層,不同隱藏層具有不同的權重;輸出層根據(jù)隱藏層權重和自身偏置輸出模型最終結果。不同層由若干神經(jīng)元組成,神經(jīng)元與神經(jīng)元間的連線為對應權重。對于單個神經(jīng)元,其輸出結果由輸入與對應權重的非線性激活函數(shù)確定[24],其表達式為
式中y為神經(jīng)元的輸出,x為神經(jīng)元的輸入,w為對應權重,為神經(jīng)元的激活閾值。
參照多元回歸模型,本研究中人工神經(jīng)網(wǎng)絡模型同樣以泥炭土、蛭石和珍珠巖3種材料在水吸力為時VG模型中相應的體積含水率和不同基質(zhì)材料的配比為自變量,復配基質(zhì)體積含水率為因變量,最終建立人工神經(jīng)網(wǎng)絡模型。
本研究選取決定系數(shù)(2)、歸一化均方根誤差(NRMSE)和平均誤差(MR)對模型預測精度進行評價,其計算式參見文獻[25-26]。
基于不同水吸力下的含水率數(shù)據(jù),利用VG模型對3種基質(zhì)材料(泥炭土、蛭石、珍珠巖)水分特征曲線進行參數(shù)擬合(表2)。由表2可知,VG模型可以很好地擬合3種基質(zhì)材料的水分特征曲線(2≥0.990)。對比3種基質(zhì)材料的水分特征曲線,3種基質(zhì)材料的飽和含水率由高到低分別為泥炭土(0.655 cm3/cm3)、蛭石(0.441 cm3/cm3)、珍珠巖(0.440 cm3/cm3)(表2)。在低吸力段(水吸力小于33 kPa),隨著水吸力的增加,3種基質(zhì)材料含水率均快速降低;3種基質(zhì)水分釋放量與飽和含水率之比為48.09%~85.68%。在中高吸力段(水吸力33~800 kPa),隨水吸力的增大,泥炭土與珍珠巖含水率變化較小,而蛭石仍有較多的水分釋放;泥炭土、珍珠巖水分釋放量與飽和含水率比分別為10.04%和5.30%,而蛭石為41.21%(圖1)。盆栽基質(zhì)的釋水特性與基質(zhì)本身的礦物組成與孔隙特征密切相關。蛭石為2:1型礦物,水分子主要填充于層間,部分水分子圍繞層間陽離子形成水合絡離子,其余呈游離態(tài)[27]。隨著水吸力的增大,層間水逐漸釋放;因此在中高吸力段,蛭石含水率逐漸降低(圖1)。泥炭土較為蓬松,有機質(zhì)含量較高,陽離子較多,存在較多結合水[28];因此在中高吸力段含水率較高。珍珠巖具有良好的透氣性,但持水性較差;水分主要附著于顆粒表面,多為重力水[10]。在釋水過程中,隨著水吸力的增大,珍珠巖含水率迅速降低;水分釋放主要發(fā)生在低吸力段。
表2 3種基質(zhì)材料的水分特征曲線參數(shù)
圖1 3種基質(zhì)材料的水分特征曲線
孔隙分布特征與基質(zhì)水分特征曲線、持水能力密切相關[29]。在水分特征曲線中,低吸力段,水分釋放主要發(fā)生在通氣孔隙中;隨著水吸力的增大,水分釋放逐漸轉(zhuǎn)變?yōu)橛行Э紫禰6,20]。泥炭土總孔隙度為65.64%,蛭石為44.08%,珍珠巖為44.03%。泥炭土和珍珠巖以>9×10-3mm孔隙為主,分別占其總孔隙的67.83%和85.68%,因此泥炭土和珍珠巖在低吸力段含水率快速降低。而二者2×10-4~9×10-3孔隙較少,分別占總孔隙的10.87%和5.30%,是其在中高吸力段水分特征曲線趨于平緩的主要原因。蛭石以>9×10-3mm和2×10-4~9×10-3mm孔隙為主,分別占其總孔隙的48.09%和43.35%。蛭石2×10-4~9×10-3mm孔隙含量較高是其在中高吸力段含水率隨水吸力的增加而降低的主要原因。泥炭土較蛭石和珍珠巖具有更多的<2×10-4mm孔隙,使得泥炭土殘余含水率(0.114 cm3/cm3)大于珍珠巖(0.039 cm3/cm3)和蛭石(0.008 cm3/cm3)(圖2)。
圖2 3種基質(zhì)材料不同當量孔徑的孔隙度
2.2.1模型構建
基于3種基質(zhì)的水分特征曲線及不同配比基質(zhì)在不同水吸力下的含水率,分別利用多元回歸和人工神經(jīng)網(wǎng)絡方法建立了不同配比基質(zhì)水分特征曲線預測模型(圖3)。對比不同建模方法水分特征曲線預測值與實測值,不同建模方法間,模型預測精度的差異主要體現(xiàn)在泥炭土-蛭石和泥炭土-珍珠巖2類復配基質(zhì)。人工神經(jīng)網(wǎng)絡模型對泥炭土-蛭石復配基質(zhì)水分特征曲線的預測精度優(yōu)于多元回歸模型;對于泥炭土‐珍珠巖復配基質(zhì),結果則相反;但二者精度差異不顯著(表3)。在不同水吸力條件下,相較于人工神經(jīng)網(wǎng)絡,多元回歸模型在5個處理中低估了復配基質(zhì)的含水率(圖3)。這主要是由于多元回歸模型低估了復配基質(zhì)的飽和含水率(表4),進而降低了水分特征曲線初始值[6]。
表3 不同模型對各處理不同水吸力下體積含水率的預測精度
注:NRMSE為歸一化均方根誤差;MR為平均誤差
Note:NRMSEis normalized root mean square error;MRis average error.
在中高吸力段(>33 kPa),人工神經(jīng)網(wǎng)絡低估了C5處理不同水吸力下的含水率,其原因主要是對殘余含水率的低估(圖3b,表4)。在水分特征曲線中,隨著水吸力的增大,基質(zhì)中通氣孔隙中的水分最先釋放,其次為有效孔隙,最后為無效孔隙[6]。θ主要受無效孔隙(<2×10-4mm)的影響,進而決定高吸力段水分特征曲線的形狀[29]。多元回歸模型對C6與C7處理在中吸力段(33~100 kPa)含水率存在明顯的高估(圖3c和圖3d)。對比實測與預測水分特征曲線VG模型參數(shù),發(fā)現(xiàn)多元回歸模型低估了參數(shù)(表4),導致C6和C7處理水分特征曲線在中吸力段較實測更平緩[30]。對于C8處理,多元回歸模型對θ的低估是導致其在初始階段預測值偏低的主要原因(圖3e,表4),同時對參數(shù)的低估是導致其低吸力段(<33 kPa)預測值偏大的主要原因。
圖3 不同處理實測和預測水分特征曲線的對比
表4 水分特征曲線參數(shù)實測值與預測值的對比
對于將泥炭土與珍珠巖作為復配材料的處理,多元回歸模型對水分特征曲線預測的預測精度要高于人工神經(jīng)網(wǎng)絡模型(圖3f~圖3h,表3)。在低吸力段(<33 kPa),人工神經(jīng)網(wǎng)絡模型對C10與C11處理含水率存在明顯的高估,其原因主要是由于該模型對參數(shù)的預測存在較大的誤差(圖3g,圖3h,表4)。在VG模型中,為進氣值倒數(shù),反映土壤初始排水時難易程度;越小,代表土壤初始排水越難,水分特征曲線在初始階段越平緩[31]。
2.2.2 不同建模方法的穩(wěn)定性分析
模型的最佳精度不是評判模型優(yōu)劣的唯一標準,還應考慮模型的穩(wěn)定性[32-34]。模型的精度主要與建模和驗證過程所選取的數(shù)據(jù)集有關。為了消除特定數(shù)據(jù)集(某一次選取的建模數(shù)據(jù)與驗證數(shù)據(jù))對模型精度造成的隨機偏差[21],本研究通過bootstrap方法進行了30次數(shù)據(jù)隨機分組、建模和驗證,用以綜合評價不同建模方法所建立模型的預測精度和穩(wěn)定性。結果表明,在30次建模和驗證過程中,人工神經(jīng)網(wǎng)絡模型實現(xiàn)的最佳精度要優(yōu)于多元回歸,但其NRMSE、2和MR的方差與分布范圍均大于多元回歸(圖4)。此外,建模和驗證過程中多元回歸模型的NRMSE的均值小于人工神經(jīng)網(wǎng)絡模型,2的均值大于人工神經(jīng)網(wǎng)絡模型,且MR更接近于0。因此,在本研究中,多元回歸建立的預測模型較人工神經(jīng)網(wǎng)絡具有更好的穩(wěn)定性[21]。此外,人工神經(jīng)網(wǎng)絡為黑箱模型,無法直觀解釋其中非線性函數(shù)的實際物理意義;且隨著人工神經(jīng)網(wǎng)絡模型規(guī)模的增大,其非線性函數(shù)呈指數(shù)增加,使得模型更為復雜,不便于使用[35]。相比之下,多元回歸模型的表達形式具體且直觀,模型的易用性更好[22]。因此,多元回歸方法是建立基質(zhì)水分特征曲線預測模型的最優(yōu)方法,本研究中基于多元回歸方法建立的最優(yōu)基質(zhì)水分特征曲線預測模型的表達式為
式中θ為水吸力S時的基質(zhì)體積含水率,cm3/cm3;S為水吸力,cm (1 cm=0.098 kPa);Vpea、Vver、Vper分別為不同基質(zhì)中泥炭土、蛭石與珍珠巖的體積百分數(shù),%。
為驗證多元回歸模型對不同配比基質(zhì)水分特征曲線的預測精度,本研究重新配置了6種不同配比基質(zhì)(表5),測定其水分特征曲線,與模型模擬結果(將水和單位換算為kPa)進行對比,結果如圖5所示。
由圖5可知,多元回歸模型對不同配比基質(zhì)水分特征曲線具有較高的預測精度(2≥0.950,較小的NRMSE,MR接近于0)(圖5)。對比不同組合復配基質(zhì),多元回歸模型對泥炭土‐珍珠巖復配基質(zhì)較泥炭土‐蛭石具有更高的預測精度。對于泥炭土‐蛭石‐珍珠巖復配基質(zhì),多元回歸模型對不同水吸力的含水率預測精度存在一定差異,但整體上仍具有較高的預測精度(2>0.950)??梢姡诙嘣貧w方法建立的基質(zhì)水分特征曲線預測模型具有可靠性。
表5 用于模型驗證的基質(zhì)材料配比體積比
圖5 不同處理實測與多元回歸模型模擬水分特征曲線的對比
為了快速獲取不同研究中盆栽基質(zhì)的水分特征曲線,本研究以盆栽控水試驗常用的泥炭土、蛭石和珍珠巖為基質(zhì)材料,測定了不同配比基質(zhì)的水分特征曲線,通過不同方法(多元回歸模型、人工神經(jīng)網(wǎng)絡)建立了其預測模型。結果表明:
1)van Genuchten模型對3種基質(zhì)水分特征曲線均具有較高的擬合精度(2≥0.990);
2)人工神經(jīng)網(wǎng)絡和多元回歸模型對不同配比基質(zhì)水分特征曲線的預測精度差異不大,但人工神經(jīng)網(wǎng)絡模型在不同配比基質(zhì)水分特征曲線的預測穩(wěn)定性低于多元回歸模型;
3)綜合考慮模型的精度和穩(wěn)定性,多元回歸方法是建立不同配比基質(zhì)水分特征曲線預測模型的最佳方法,經(jīng)驗證,其精度高于0.9,基于該方法建立的模型為不同配比盆栽基質(zhì)水分特征曲線的快速獲取提供了可靠途徑。
[1] 宋松泉,王彥榮. 植物對干旱脅迫的分子反應[J]. 應用生態(tài)學報,2002,13(8):1037-1044.
SONG Shongquan, WANG Yanrong. Molecular response of plant to drought stress[J]. Chinses Journal of Applied Ecology, 2002, 13(8): 1037-1044. (in Chinese with English abstract)
[2] 王衛(wèi)星,楊明欣,高鵬,等. 基于多光譜和氣象參數(shù)的菜心水分脅迫指數(shù)反演[J]. 農(nóng)業(yè)工程學報,2022,38(6):157-164.
WANG Weixing, YANG Mingxin, GAO Peng, et al. Inverting the water stress index of theusing multiple-spectral and meteorological parameters[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2022, 38(6): 157-164. (in Chinese with English abstract)
[3] LI G, LIN L, DONG Y Y, et al. Testing two models for the estimation of leaf stomatal conductance in four greenhouse crops cucumber, chrysanthemum, tulip and lilium[J]. Agricultural and Forest Meteorology, 2012, 165: 92-103.
[4] ZHANG N Y, LI G, YU S X, et al. Can the responses of photosynthesis and stomatal conductance to water and nitrogen stress combinations be modeled using a single set of parameters?[J]. Frontiers in Plant Science, 2017, 8: 328.
[5] CAI G C, van DERBORGHT J, COUVREUR V, et al. Parameterization of root water uptake models considering dynamic root distributions and water uptake compensation[J]. Vadose Zone Journal, 2018, 17(1): 160125.
[6] 邵明安,王全九,黃明斌. 土壤物理學[M]. 北京:高等教育出版社,2006.
[7] LAMHAMEDI M S, MARGOLIS H, RENAUD M, et al. Effects of different irrigation intensities on growth, mineral nutrition and nutrient leaching of 1+0 black spruce produced in air-slit containers in a forest nursery[J]. Canadian Journal of Forest Research, 2003, 33(2): 279-291.
[8] CHAMBEL M R, CLIMENT J, ALIA R. Divergence among species and populations of Mediterranean pines in biomass allocation of seedlings grown under two watering regimes[J]. Annals of Forest Science, 2007, 64(1): 87-97.
[9] CARLES S, LAMHAMEDI M S, STOWE D C, et al. Frost tolerance of two-year-oldseedlings grown under different irrigation regimes in a forest nursery[J]. Scandinavian Journal of Forest Research, 2008, 23(2): 137-147.
[10] 朱建龍,張瀟丹,頡建明,等. 日光溫室栽培基質(zhì)有效導熱系數(shù)預測模型[J]. 農(nóng)業(yè)工程學報,2021,37(24):199-207.
ZHU Jianlong, ZHANG Xiaodan, XIE Jianming, et al. Prediction models for the effective thermal conductivity of cultivation substrates in solar greenhouses[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2021, 37(24): 199-207. (in Chinese with English abstract)
[11] 張乃于,閆雙堆,任倩,等. 珍珠巖粒徑對土壤水分運移的影響[J]. 灌溉排水學報,2019,38(2):22-28.
ZHANG Naiyu, YAN Shuangdui, REN Qian, et al. Water movement in soil amended with perlite particles of different sizes[J]. Journal of Irrigation and Drainage, 2019, 38(2): 22-28. (in Chinese with English abstract)
[12] JOHN A, FUENTES H R, GEORGE F. Characterization of the water retention curves of Everglades wetland soils[J]. Geoderma, 2021, 381: 114724.
[13] 張雅馥,王金滿,王敬朋,等. 生物炭添加對礦區(qū)壓實土壤水力特性的影響[J]. 農(nóng)業(yè)工程學報,2021,37(22):58-65.
ZHANG Yafu, WANG Jinman, WANG Jingpeng, et al. Effects of biochar addition on the hydraulic properties of compacted soils in mining areas[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2021, 37(22): 58-65. (in Chinese with English abstract)
[14] 鄧羽松,丁樹文,蔡崇法,等. 鄂東南崩崗剖面土壤水分特征曲線及模擬[J]. 土壤學報,2016,53(2):355-364.
DENG Yushong, DING Shuwen, CAI Congfa, et al. Characteristic curves and model analysis of soil moisture in collapse mound profiles in southeast Hubei[J]. Acta Pedologica Sinica, 2016, 53(2): 355-364. (in Chinese with English abstract)
[15] RAM S, PRASAD K S H, GAIROLA A, et al. Estimation of border-strip soil hydraulic parameters[J]. Journal of Irrigation and Drainage Engineering, 2012, 138(6): 493-502.
[16] QIAO J B, ZHU Y J, JIA X X, et al. Development of pedotransfer functions for soil hydraulic properties in the critical zone on the Loess Plateau, China[J]. Hydrological Processes, 2018, 32(18): 2915-2921.
[17] BAI X, SHAO M A, JIA X X, et al. Prediction of the van Genuchten model soil hydraulic parameters for the 5-m soil profile in China’s Loess Plateau[J]. Catena, 2022, 210: 105889.
[18] van GENUCHTEN M T. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils[J]. Soil Science Society of America Journal, 1980, 44: 892-898.
[19] 盧佳宇,王輝,歐陽贊. 低鹽再生水灌溉對亞熱帶紅壤水力特性及微觀結構的影響[J]. 農(nóng)業(yè)工程學報,2022,38(18):103-112.
LU Jiayu, WANG Hui, OUYANG Zan. Effects of low-salinity reclaimed wastewater irrigation on the hydraulic properties and microstructure of subtropical red soil[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2022, 38(18): 103-112. (in Chinese with English abstract)
[20] 雷志棟,楊詩秀,謝森傳. 土壤水動力學[M]. 北京:清華大學出版社,1988.
[21] ZHAO C L, SHAO M A, JIA X X, et al. Using pedotransfer functions to estimate soil hydraulic conductivity in the Loess Plateau of China[J]. Catena, 2016, 143: 1-6.
[22] WANG Y Q, SHAO M A, LIU Z P, et al. Prediction of bulk density of soils in the Loess Plateau Region of China[J]. Surveys in Geophysics, 2014, 35(2): 395-413.
[23] 馮正江,聶衛(wèi)波,余淼,等. 多尺度土壤入滲特性的變異特征和傳遞函數(shù)構建[J]. 農(nóng)業(yè)工程學報,2022,38(13):64-75.
FENG Zhengjiang, NIE Weibo, YU Miao, et al. Multiple scale variability of soil infiltration characteristics and establishment of pedo-transfer function[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2022, 38(13): 64-75. (in Chinese with English abstract)
[24] 焦李成,楊淑媛,劉芳,等. 神經(jīng)網(wǎng)絡七十年:回顧與展望[J]. 計算機學報,2016,39(8):1697-1716.
JIAO Licheng, YANG Shuyuan, LIU Fang, et al. Seventy years beyond neural networks: Retrospect and prospect[J]. Chinese Journal of Computers, 2016, 39(8): 1697-1716. (in Chinese with English abstract)
[25] 焦懷瑾,陳崇成,黃洪宇. 結合ICESat-2和GEDI的中國東南丘陵地區(qū)ASTERGDEM高程精度評價與修正[J]. 地球信息科學學報,2023,25(2):409-420.
JIAO Huaijin, CHEN Chongcheng, HUANG Hongyu. Elevation accuracy evaluation and correction of ASTER GDEM in China southeast hilly region by combining ICESat-2 and GEDI data[J]. Journal of Geo-Information Science, 2023, 25(2): 409-420. (in Chinese with English abstract)
[26] 萬華,高曉東,趙西寧,等. 黃土區(qū)蘋果園土壤剖面水、汽通量模擬研究[J]. 土壤學報,2022,60(3):1-14.
WANG Hua, GAO Xiaodong, ZHAO Xining, et al. Simulation of water and vapor fluxes in soil profile of apple orchard in Loess area[J]. Acta Pedologica Sinica, 2022, 60(3): 1-14. (in Chinese with English abstract)
[27] SU X L, MA L Y, WEI J M, et al. Structure and thermal stability of organo-vermiculite[J]. Applied Clay Science, 2016(132/133): 261-266.
[28] 鄭旭,張宗儉,段韞丹,等. 濕潤劑對基質(zhì)水分吸持和白菜穴盤苗生長的影響[J]. 農(nóng)業(yè)工程學報,2023,39(1):188-194.
ZHENG Xu, ZHANG Zongjian, DUAN Yundan, et al. Effects of wetting agent application on substrate water absorption and Chinese cabbage plug seedlings[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2023, 39(1): 188-194. (in Chinese with English abstract)
[29] XING X G, KANG D G, MA X Y. Differences in loam water retention and shrinkage behavior: Effects of various types and concentrations of salt ions[J]. Soil & Tillage Research, 2017, 167: 61-72.
[30] 董義陽,趙成義,于志同,等. 古爾班通古特沙漠南緣丘間地土壤水分特征曲線及模擬[J]. 水土保持學報,2017,31(1): 166-171.
DONG Yiyang, ZHAO Chengyi, YU Zhitong, et al. Characteristic curves and models analysis of soil water in interdune at the southern edge of Gurbantunggut Desert[J]. Journal of Soil and Water Conservation, 2017, 31(1): 166-171. (in Chinese with English abstract)
[31] SCHAAP M G, van GENUCHTEN M T. A modified Mualem-van Genuchten formulation for improved description of the hydraulic conductivity near saturation[J]. Vadose Zone Journal, 2006, 5(1): 27-34.
[32] CORNELIS W M, RONSYN J, van MEIRVENNE M, et al. Evaluation of pedotransfer functions for predicting the soil moisture retention curve[J]. Soil Science Society of America Journal, 2001, 65: 638-648.
[33] AGYARE W A, PARK S J, VLEK P L G. Artificial neural network estimation of saturated hydraulic conductivity[J]. Vadose Zone Journal, 2007, 6: 423-431.
[34] PASCHALIS A, BONETTI S, GUO Y R, et al. On the uncertainty induced by pedotransfer functions in terrestrial biosphere modeling[J]. Water Resources Research, 2022, 58(9): e2021WR031871.
[35] 姚立忠,李太福,易軍,等. 神經(jīng)網(wǎng)絡模型的透明化及輸入變量約簡[J]. 計算機科學,2012,39(9):247-251,278.
YAO Lizhong, LI Taifu, YI Jun, et al. Visualize black-box of NN model and its application in dimensionality reduction[J]. Chinese Journal of Computers, 2012, 39(9): 247-251, 278. (in Chinese with English abstract)
Comparative analysis and prediction of the potting media water retention curve with different proportional compositions
ZHU Zhaocen1,2, SHAO Ming′an1,2,3, ZHAO Chunlei1,3※, JIA Xiaoxu1,3, WANG Jiao1,3
(1.,,100101,; 2.,,712100,; 3.,,100190,)
The water stress threshold is one of the most important indexes to evaluate the response of crops to water stress. Water control in the pot culture is also a commonly-used way to simulate drought stress conditions. The absorption of soil water by crops is mainly decided by the soil water potential. And the relationship between soil water content and water potential varies with matric medias. However, the water stress threshold of a crop is usually given in the form of mass water content. However, there is some inconsistency in the matric media under previous pot experiments. This study aims to rapidly obtain the water retention curves of base materials in the pot experiments, in order to predict the potting soil water retention curves. The peat soil, vermiculite, and perlite were taken as matric materials in the pot experiments. The water retention curves of various media were measured, including three kinds of matric materials and eight different proportional compositions. The van Genuchten model was used to fit these water retention curves. After that, the pedotransfer functions (PTFs) of water retention curves were established using multiple regression and artificial neural network. The normalized root mean square error (NRMSE), determination coefficient (2), and mean error (MR) were selected as the accuracy indicators of the model. Furthermore, the model stability was verified through 30 cycles to finally determine the optimal modeling and the predicted water retention curve. The results showed that an ideal fitting effect of the van Genuchten model was achieved in the water retention curves of (2>0.99). There were some significant differences in the water retention curves of the three potting medias. Peat soil presented more invalid pores than vermiculite and perlite, resulting in a large water content in the high-water suction section (water suction>100 kPa). The high capillary pore content of vermiculite led to a high-water content in the middle and high-water suction sections (water suction 33-800 kPa). Pearlite shared the higher permeability and low water holding capacity. Water was rapidly released with the increase of water suction. The prediction accuracies of the artificial neural network model for the peat soil water retention curves and vermiculite media materials were better than the multiple regression model. However, the opposite trend was observed for the peat soil and perlite media materials. Two different prediction models had different errors for the water retention curve parameters in the different media materials combinations. The multiple regression models often underestimated the prediction of saturated water content. As such, the multiple regression models underestimated the water content of media materials, compared with artificial neural networks under different water suction. The artificial neural network with PTFs reached a higher accuracy than the multiple regression with PTFs. The accuracy of multiple regression with PTFs had smaller variation and higher stability (smallerNRMSE, larger2, and mean error closer to 0), compared with the artificial neural network. Meanwhile, there was a large variance and distribution range ofNRMSE,2, andMRof PTFs developed by artificial neural networks, compared with the multiple regression. Therefore, better stability was achieved in the PTFs developed by multiple regression. In terms of accuracy and stability, multiple regression can be selected as the best to develop the water retention curve PTFs in various plot experiment medias. The finding can provide a strong reference for the rapid acquisition of water retention curves using the comparison between the pot water control experiments.
models; media materials; water stress; potting experiment; water retention curve; pedotransfer functions
2023-01-09
2023-02-28
中國科學院A類戰(zhàn)略性先導科技專項資助(XDA28130100);國家自然科學基金項目(41907009);黃土高原土壤侵蝕與旱地農(nóng)業(yè)國家重點實驗室開放基金項目(A314021402–2014)
朱釗岑,博士生,研究方向為土壤物理和水文生態(tài)。Email:zhuzhaocen@nwafu.edu.cn
趙春雷,博士,副研究員,碩士生導師,研究方向為土壤物理和農(nóng)業(yè)生態(tài)。Email:zhaocl@igsnrr.ac.cn
10.11975/j.issn.1002-6819.202301037
S152; S220.4
A
1002-6819(2023)-08-0197-08
朱釗岑,邵明安,趙春雷,等. 不同盆栽基質(zhì)水分特征曲線的對比分析與模擬[J]. 農(nóng)業(yè)工程學報,2023,39(8):197-204. doi:10.11975/j.issn.1002-6819.202301037 http://www.tcsae.org
ZHU Zhaocen, SHAO Ming′an, ZHAO Chunlei, et al. Comparative analysis and prediction of the potting media water retention curve with different proportional compositions[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2023, 39(8): 197-204. (in Chinese with English abstract) doi:10.11975/j.issn.1002-6819.202301037 http://www.tcsae.org