Clarifying Spatial-Temporal Variability of Surface Soil Salinization in Arid Cotton Fields
-
摘要: 明确土壤盐渍化时空变异特征是确保干旱区棉田精准灌溉和作物良好生长发育的基础。考虑棉田不同灌溉节点和灌溉方式,利用大地电导率仪(EM38-MK2),在棉花生育期内进行3次表观电导率数据和土样的采集,采用局部建模和全局建模的思路构建表观电导率与实测电导率间的反演模型。综合利用经典统计方法和地统计学方法,分析表层土壤盐渍化的时空变异特征。结果表明,表观电导率与实测电导率建立的局部模型具有较好精度,R2均大于0.79,而全局模型的R2仅为0.52,表明局部模型优于全局模型;不同时期表层电导率变异性差异较大,3月和10月变异系数均小于50%,属中等变异,而7月变异系数大于50%,属强变异;各时期半变异函数的最优模型均为球状模型,基台值与块金值之比均小于25%,表明表层盐分变异主要由气候,蒸降比、地下水埋深度等结构性因素引起;研究区盐渍化土壤面积由播种前的0.49%增加到棉花收获后的98.23%,在棉花生育期内,表层土壤以轻度盐渍化和中度盐渍化为主;土壤盐渍化空间变异强度主要以中等变异为主,强变异次之,弱变异所占面积最小,而强变异集中分布在研究区南端,弱变异区域主要集中在研究区中部。研究结果为干旱区田间尺度的棉田土壤盐渍化时空变异研究提供了一定的思路和方法。Abstract: Clarifying the spatial-temporal variation characteristics of soil salinization is the basis for ensuring precision irrigation and crop growth in arid cotton fields. Considering different irrigation nodes and modes in cotton fields, apparent electrical conductivity was measured and soil samples were collected three times during the growth period of cotton by using the electromagnetic senser (EM38-MK2). The inversion model between apparent electrical conductivity and measured electrical conductivity was constructed. The spatial-temporal variation characteristics of surface soil salinization were analyzed by using classical statistical methods and geostatistical methods. The results showed that the local model established by the apparent electrical conductivity and the measured electrical conductivity had a good accuracy with a R2 greater than 0.79, which was superior to the global model with a R2 only 0.52, indicating that the local model is superior to the global model. The surface soil electrical conductivity variability was significantly different in different periods. The coefficient of variation in March and October was less than 50% (a medium variation), while that in July was more than 50% (a strong variation). The optimal semi-variogram models in each period were spherical, and the ratio of sill to nugget value was less than 25%, indicating that the variation of surface salinity was mainly caused by structural factors such as climate, evaporation ratio and groundwater. The salinized soil area in the studied area increased from 0.49% before planting to 98.23% after cotton harvest. During the growth period of cotton, the surface soil was mainly salinized slightly and moderately. The spatial variation intensity of soil salinization was mainly medium, followed by strong. The area occupied by weak variation was the smallest and was mainly concentrated in the middle part of the area, while the strong variation was concentrated in the southern part. This results provided some certain ideas and methods for studying the temporal and spatial variability of soil salinization at the field scale in arid areas.
-
Key words:
- Electromagnetic induction /
- Soil salinity /
- Salinization /
- Spatial-temporal variability
-
表 1 局部与全局电导率反演模型
Table 1. Local and global conductivity inversion models
建模方法
Method日期
Date模型
Model决定系数
R2局部 2018.03.31 Y = 0.0119X1 − 0.0037X2 + 0.4289 0.88 2018.07.07 Y = 0.0483X1 − 0.0303X2 + 0.7368 0.77 2018.10.27 Y = 0.1707X1 − 0.1216X2 + 0.8338 0.83 全局 Y = 0.0445X1 − 0.0234X2 + 1.7778 0.38 注:自变量ECv0.75与ECv1.5分别表示为X1与X2 表 2 局部模型与全局模型评价指标对比
Table 2. Comparison of evaluation indices between local model and global model
建模方法
Method日期
Date决定系数
R2均方根误差
RMSE平均绝对百分误差
MAPE平均误差
ME局部 2018.03.31 0.89 0.31 0.03 −0.11 2018.07.07 0.90 1.17 0.42 0.27 2018.10.27 0.79 1.08 0.13 0.11 全局 0.52 2.29 1.02 −0.22 表 3 不同时期表层土壤电导率统计特征值
Table 3. Statistical characteristic values of surface soil conductivity in different periods
日期
Date最小值(dS m−1)
Minimum最大值(dS m−1)
Maximum平均(dS m−1)
Mean标准差
SD变异系数(%)
CV峰度
Kurtosis偏度
Skewness2018.03.31 0.15 2.11 0.61 0.28 45.63 2.54 1.62 2018.07.07 0.36 7.71 2.64 1.37 51.98 −0.20 0.81 2018.10.27 1.49 16.59 5.38 1.76 32.76 3.39 0.69 表 4 不同时期半变异函数模型
Table 4. Semi-variogram models at different periods
日期
Date模型
Model块金值C0
Nugget(dS m−1)基台值C0 + C
Sill(dS m−1)块金值/基台值
Nugget/Sill(%)变程
Range(m)决定系数
R2残差
RSS2018.03.31 S 0.01 0.09 11.07 125.10 0.89 4.01 × 10−4 2018.07.07 S 0.34 2.10 16.30 124.90 0.89 6.59 × 10−3 2018.10.27 S 0.54 3.35 16.26 119.00 0.80 8.63 × 10−3 注:S表示球状模型 -
[1] 杨劲松. 中国盐渍土研究的发展历程与展望[J]. 土壤学报, 2008, (5): 837 − 845. [2] 彭 杰. 荒漠土壤盐渍化遥感监测与开垦方案分析[D]. 浙江杭州: 浙江大学, 2019. [3] 赵文举, 唐学芬, 范严伟, 等. 压砂地土壤盐分演替特征研究[J]. 灌溉排水学报, 2017, 36(8): 25 − 31. [4] Liang L C, Liu W T, Zhang X, et al. Research progress in phytoextraction of heavy metal contaminated saline soil[J]. Journal of Agro-Environment Science, 2016, 35(7): 1233 − 1241. [5] 窦 旭, 史海滨, 苗庆丰, 等. 盐渍化灌区土壤水盐时空变异特征分析及地下水埋深对盐分的影响[J]. 水土保持学报, 2019, 33(3): 246 − 253. [6] 李 楠. 岷江干热河谷土壤养分空间异质性研究[D]. 四川成都: 四川师范大学, 2016. [7] 姚荣江, 杨劲松, 邹 平, 等. 基于电磁感应仪的田间土壤盐渍度及其空间分布定量评估[J]. 中国农业科学, 2008, (2): 460 − 469. [8] 李洪义, 史 舟, 吴次芳, 等. 基于EM38的滨海盐土剖面电导率原位测定[J]. 土壤学报, 2013, 50(6): 1231 − 1235. [9] 刘新路, 彭 杰, 冯春晖, 等. 基于电磁感应仪数据的南疆棉田土壤电导率反演模型研究[J]. 土壤学报, 2020, 57(3): 646 − 655. [10] 吴家林, 彭 杰, 冯春晖, 等. 基于电磁感式大地电导率仪的南疆干旱区土壤电导率反演研究[J]. 土壤通报, 2019, 50(6): 1278 − 1284. [11] 李 兵, 刘广明, 杨劲松, 等. 基于磁感式大地电导率仪的土壤盐分解译模型[J]. 土壤, 2017, 49(4): 789 − 794. [12] Rhoades J D, Ingvalson R D. Determining salinity in field soils with soil resistance measurements[J]. Soil Science Society of America Journal, 1971, 35(1): 54 − 60. doi: 10.2136/sssaj1971.03615995003500010020x [13] Rhoades J D, Van Schilfgaarde J. An electrical conductivity probe for determining soil salinity[J]. Soil Science Society of America Journal, 1976, 40(5): 647 − 651. doi: 10.2136/sssaj1976.03615995004000050016x [14] Rhoades J D, Lesch S M, Lemert R D, et al. Assessing irrigation/drainage/salinity management using spatially referenced salinity measurements[J]. Agricultural Water Management, 1997, 35(1-2): 147 − 165. doi: 10.1016/S0378-3774(97)00017-6 [15] Yao R J, Yang J S, Liu G M. Calibration of soil electromagnetic conductivity in inverted salinity profiles with an integration method[J]. Pedosphere, 2007, 17(2): 246 − 256. doi: 10.1016/S1002-0160(07)60031-9 [16] Jiang Q S, Peng J, Biswas A, et al. Characterising dryland salinity in three dimensions[J]. Science of The Total Environment, 2019, 682: 190 − 199. doi: 10.1016/j.scitotenv.2019.05.037 [17] 吴亚坤, 刘广明, 杨劲松, 等. 基于电磁感应研究新疆土壤盐分三维空间变异对季节的响应[J]. 农业工程学报, 2017, 33(14): 173 − 178. [18] 邓 凯, 丁建丽, 杨爱霞, 等. 基于电磁感应技术的土壤剖面盐分空间分布建模研究[J]. 生态学报, 2016, 36(20): 6387 − 6396. [19] 宋江辉, 朱永琪, 陈建华, 等. 基于电磁感应仪的土壤盐渍化剖面特征解译研究[J]. 土壤通报, 2017, 48(3): 552 − 559. [20] Khongnawang T, Zare E, Srihabun P, et al. Comparing electromagnetic induction instruments to map soil salinity in two-dimensional cross-sections along the Kham-rean Canal using EM inversion software[J]. Geoderma, 2020, doi: org/10.1016/j.geoderma.2020.114611. [21] 孙运朋, 陈小兵, 张振华, 等. 滨海棉田土壤盐分时空分布特征研究[J]. 土壤学报, 2013, 50(5): 891 − 899. [22] 付腾飞, 张 颖, 徐兴永, 等. 山东滨海低平原区盐渍土盐分的时空变异研究[J]. 海洋开发与管理, 2017, 34(12): 38 − 45. [23] 刘广明, 吕真真, 杨劲松, 等. 典型绿洲区土壤盐分的空间变异特征[J]. 农业工程学报, 2012, 28(16): 100 − 107. [24] 贡 璐, 张雪妮, 吕光辉, 等. 塔里木河上游典型绿洲不同土地利用方式下土壤质量评价[J]. 资源科学, 2012, 34(1): 120 − 127. [25] 牛建龙, 彭 杰, 王家强, 等. 新疆阿拉尔地区近53年气候变化特征分析[J]. 干旱区资源与环境, 2016, 30(1): 72 − 77. [26] 王建勋, 胡云喜, 朱建雯. 1961-2000年新疆阿拉尔垦区蒸发量的变化特征[J]. 气象与环境学报, 2008, 24(5): 29 − 31. [27] 班春广, 陶 辉, 董义阳, 等. 未来气候情景下塔里木河干流日径流过程模拟研究[J]. 干旱区研究, 2018, 35(4): 770 − 778. [28] 鲍士旦. 土壤农化分析[M]. 北京: 中国农业出版社. 2016, 180−182. [29] 刘 宁. 不同土地利用方式下黄河三角洲土壤特性空间变异研究[D]. 山东: 山东农业大学, 2007. [30] 臧 亮, 张贵军, 张 慧, 等. 黄骅市土壤重金属空间变异特征及污染评价[J]. 水土保持研究, 2017, 24(5): 337 − 342. [31] 汪 花, 刘秀明, 刘 方, 等. 喀斯特地区小尺度农业土壤砷的空间分布及污染评价[J]. 环境科学, 2019, 40(6): 2895 − 2903. [32] Staff U S L. USDA Handbook No 60: Diagnosis and improvement of saline and alkali soils[M]. USDA, Washington, DC, 1954.