稀疏格网样点的国家尺度土壤属性制图方法与应用

 国家尺度土壤属性数据是地球生物化学循环及水循环等领域研究的重要数据,目前,该尺度土壤属性数据的获取方法主要有两类:土壤属性-空间数据连接法和空间插值。为了确定哪一类方法更适合稀疏样点的国家尺度土壤属性制图,本文以中国吉林省的土壤有机质含量制图为例,采用8~32km格网样点和1∶100万土壤图,对这两类方法进行对比分析。独立样本验证结果表明,土壤属性-空间数据连接法的平均误差(ME)大于距离反比加权(IDW)插值,而平均绝对误差(MAE)和均方根误差(RMSE)都小于IDW插值。IDW插值获得的土壤属性图虽然能大致反映土壤属性空间分布的基本规律,但出现了类似"牛眼睛"的空间结构,且存在无样点区估计值不准确等问题;土壤属性-空间数据连接法尽管忽略了同种土壤类型内部的差异,保留了不同土壤类型边界处的属性值突变,但获得的土壤属性图更能反映土壤属性分布的基本规律,也具有比较详细的土壤属性空间结构。因此,在基于稀疏样点的国家尺度土壤属性制图中,土壤属性-空间数据连接法的制图效果要优于IDW空间插值法。     

土地利用变化对吴江市水田土壤有机碳储量的影响分析

 农业表层土壤碳库容易受人为强烈干扰,而又可以在较短的时间尺度上进行调节,当今我国经济发达地区土地利用变化必然会对土壤固碳产生重要影响。本研究以江苏省吴江市水稻土为例,利用新一代中分辨率成像光谱仪(MODIS)和TM/ETM影像提取了1984年稻田面积,以及这部分稻田在2000-2005年的土地利用变化状况。研究中以最大似然法对TM/ETM、MODIS影像应用归一化植被指数(NDVI)、增强型植被指数(EVI)和陆地水分指数(LSWI)掩膜的方法作了识别提取;同时,结合第二次全国土壤普查、2003年耕地地力调查点和吴江市农林局土肥指导站长期定位点的土壤有机碳数据估算了1984年和2000-2005年土壤碳库变化情况。结果表明:近20多年来尽管吴江市水稻土水耕熟化过程中有机碳总体呈增加的趋势,但由于大量稻田被非农用地所取代,导致土壤固碳能力大幅度下降,尤其从2001年开始从"碳汇"变成"碳源"。因此,在我国经济发达区应密切关注耕地转换成非农用地而导致的土壤有机碳的损失。     

商河县耕层土壤有机碳密度及储量研究

采用商河县2010年耕地地力评价有机质采样点的数据,运用GIS和地统计学相结合的方法,研究了商河县土壤有机质的密度及储量,并按照土类统计分析了不同土壤类型间的有机碳密度及储量差异。研究表明,商河县土壤有机碳密度范围为1.69~5.17kg/m2,平均3.44kg/m2;有机碳储量为0.26×1010kg。按乡镇统计,有机碳密度最大值为玉皇庙镇3.65kg/m2,最小值为郑陆镇3.15kg/m2,有机碳储量最大值玉皇庙镇3.47×108 kg,最小值张坊乡0.92×108 kg;按土壤类型统计,有机碳密度最大的为湿潮土3.46kg/m2,最小为风沙土3.29kg/m2,有机碳储量最大为潮土10.38×108 kg,最小为褐土化潮土1.34×108 kg。     

基于CEEMDAN-LSTM组合方法的海平面变化预测分析北大核心CSCD

为提高海平面变化预测精度,将自适应噪声完备集合经验模态分解(CEEMDAN)与长短期记忆网络(LSTM)相结合,采用“分解-预测-重组”思路,提出一种海平面变化预测组合方法CEEMDAN-LSTM。结果表明,相对于直接使用LSTM神经网络进行预测(MAE=16.87 mm, RMSE=21.51 mm),以及已有的EEMD-BP神经网络组合方法(MAE=10.4 mm, RMSE=15.44 mm),CEEMDAN-LSTM组合方法预测表现最优(MAE=8.89 mm, RMSE=11.34 mm),具有最低的预测误差。     

基于时序影像的农业活动因子提取与闽西耕地SOC数字制图

人类活动对表层耕地土壤有机碳（Soil Organic Carbon, SOC）影响强烈,但目前大范围复杂地貌地形区的耕地SOC数字制图对人为因素的空间刻画不足。本文以福建省西部耕地为研究对象,基于Sentinel-2/MSI时间序列数据提取轮作模式分类信息（Crop Rotation, CR）,以及可反映轮作模式信息的植被特征变换变量（Harmonic Analysis of Time Series, HANTS）,分别作为农业活动定性和定量因子,将常规气候和地形因子作为自然环境因子,并对不同类型环境变量进行组合（气候+地形、气候+地形+轮作模式、气候+地形+HANTS变量、气候+地形+轮作模式+HANTS变量）。基于随机森林模型（Random Forest, RF）对不同环境变量组合驱动的耕地表层SOC空间预测精度进行对比分析,探索以轮作模式为例的农业活动因子提高耕地表层SOC数字制图精度的可能性。结果表明,同时加入两种农业活动因子的RF模型表现最佳,其模型预测精度相较于纯自然环境变量驱动的模型有明显提高（R²提高了89.47%,RMSE和MAE分别下降了10.66%和12.05%）。轮作模式类型（CR）和HANTS变量两种农业活动因子均被保留参与建模,尤其是轮作模式类型显著影响耕地SOC,在最佳模型的环境变量重要性中排序第四。由此可见,轮作模式相关农业活动因子可有效提高耕地SOC空间预测精度。在所有RF模型中,年降水量（Annual Rainfall, Rainfall）的重要性排名都是第一位。通过最佳模型反演得出该区耕地土壤有机碳均值为18.22±2.99 g/kg,范围为8.25~30.69 g/kg,双季稻和烟稻种植区域SOC含量高于稻菜种植区域。研究结果为复杂地貌地形区耕地土壤有机碳协同变量的更新提供了新的思路。     

县域尺度下湖南省碳排放空间分异特征与影响因素

开展县域尺度下的碳排放空间分异与影响因素研究,是助推实现县域等多尺度区域“双碳”目标的重要环节。以湖南省122个县(市、区)为研究对象,通过夜间灯光数据估算湖南省各区县的碳排放量,采用探索性空间数据分析方法刻画了县域碳排放时空格局,运用地理探测器和时空地理加权回归(GTWR)模型,从政策、经济、能源、社会和产业维度对县域碳排放空间分异与影响因素进行了定量研究。结果表明：(1)2012～2020年,湖南省县域碳排放总体呈减弱趋势,且差异较为明显,空间格局分布为北高南低,东西差距较小;(2)莫兰指数(Moran’s I)逐年下降,空间集聚特征较为显著,总体成正相关关系;(3)财政支出、人口规模、城镇化率、能源碳排放强度和农业发展水平是影响县域碳排放的主导因子;(4)通过时空地理加权回归模型对主导因子进行分析,发现同一指标对不同区县碳排放的影响存在显著的时空差异。     

基于Prophet-XGBoost模型的GNSS高程时间序列预测

针对GNSS时间序列非平稳性和非线性等特点,通过分析XGBoost模型与Prophet模型的适用性与特点,构建Prophet-XGBoost预测模型。该模型先通过Prophet模型对GNSS原始时间序列进行分解,然后通过XGBoost模型进行分部预测,等权相加得到预测结果。实验选用ALGO、ALRT、BRST三个IGS站U分量日坐标时间序列数据,采用MAE和RMSE作为评价指标。结果表明,与单一的XGBoost模型和Prophet模型相比,Prophet-XGBoost模型的MAE和RMSE值均得到一定程度优化,说明该模型具备有效性,可用于GNSS时间序列预测。     

干旱区剖面土壤盐分空间变异特征及随机模拟

利用便携式土壤盐分计测定干旱区膜下滴灌棉田及裸地剖面土壤电导率,利用GS+软件确定其半方差函数,并进行序
贯高斯模拟来研究土壤剖面盐分的空间变异特征。结果显示:干旱区剖面上土壤盐分空间变异性强,膜下滴灌棉田符合球形模
型,裸地区符合高斯模型,且棉田变程远小于裸地;棉田埋深50cm以浅土壤盐分变化剧烈,埋深50cm 以深变化平缓,在垂向上
呈现3个高值区(30,50和75cm深度附近)及3个低值区(埋深40,60和100cm附近);克里格法具有明显的平滑作用,降低了土
壤盐分的空间变异性,而序贯高斯模拟数据更离散,更能反映土壤盐分的空间变异性,但多次模拟均值随次数增加趋于稳定。      

多尺度视角下的青藏高原水资源短缺估算及空间格局

青藏高原水资源总量丰富,但由于水资源量与用水量在空间上分布不均衡,部分人口、城镇密集地区水资源短缺严重。本研究在多源数据的基础上,通过空间分析、降尺度处理等,建立了青藏高原省区、市域、县域空间尺度的水资源与用水量数据集。通过比较5、10、20、30年重现期多空间尺度的水资源短缺程度,分析水资源短缺在青藏高原的尺度效应,揭示青藏高原水资源短缺格局与特征,识别面临水资源短缺的人口与面积。结果表明,青藏高原在省区尺度无水资源短缺;在15个市域单元中,有3个市域出现水资源短缺;在115个县域单元中,有29个县域呈现出不同程度的水资源短缺,水资源短缺县域主要集中在青海省的河湟谷地、柴达木盆地与西藏自治区的一江两河流域等人口、城镇密集区域。总体而言,由于较大空间尺度地理单元内部各县域用水强度差异,在县域尺度面临水资源短缺的人口与面积大于市域与省区尺度面临水资源短缺的人口与面积。以县域为基本单元,发现青海省与西藏自治区30年重现期面临水资源短缺的人口占总人口的56.4%,出现水资源短缺的面积占总面积的10.4%。县域之间水资源短缺指数秩相关系数计算结果显示,省区内部各县域同时出现水资源短缺的可能性较大,而省区之间各县域同时出现水资源短缺的可能性相对随机。研究结果为制定青藏高原水资源短缺管理对策、促进区域城镇化与资源环境协调发展提供科学依据。     

克里格法的土壤水分遥感尺度转换

 尺度效应往往会制约着定量遥感反演的精度,对地学信息进行空间尺度转换是生产实践的必然要求,而常用的尺度转换模型多利用光谱数据进行差值计算,不适合升尺度和降尺度转换。由于土壤含水量数据具有区域变化量的随机性和结构性特点,本文以15m分辨率的ASTER图像像元为基本单元,采用点克里格法完成ASTER 15m至7.5m分辨率的土壤含水量数据降尺度转换,从分维数的相似程度上来看,转换结果是合理的;并利用块状克里格法对地面实测样点数据进行点到7.5m分辨率的面数据升尺度转换,将升尺度和降尺度转换结果与实测样点均值相比较,结果表明:7.5m分辨率的实测样点土壤水均值误差在1.5782-5.019之间,块状克里格法获取的升尺度土壤含水量数据与点克里格法获取的降尺度土壤含水量数据之间误差则为1.2825-5.0481,可见克里格法考虑了点与周边的关系,所获得的土壤含水量值要优于未考虑空间异质性的土壤含水量平均值。     

Spatiotemporal variation of soil organic carbon in the cultivated soil layer of dry land in the South-Western Yunnan Plateau,China

The dynamics of soil organic carbon(SOC)in cropland is one of the central issues related to both soil fertility and environmental safety. However, little information is available at county level regarding the spatiotemporal variability of SOC in the southwestern mountainous region of China. Thus, this study aimed to explore spatiotemporal changes of SOC in the cultivated soil layer of dry land in Mojiang County,Yunnan Province, China. Data were obtained from the second national soil survey(SNSS) of 1985 and soil tests for fertilizer application carried out by the Mojiang Agricultural Bureau in 2006. The ANOVA test was applied to determine any significant differences between the datasets, while semivariogram analysis was performed on geostatistics via an ordinary Kriging method in order to map spatial patterns of soil organic carbon density(SOCD). The results revealed that SOCD in the cultivated soil layer significantly decreased from 3.93 kg m~(-2) in 1985 to 2.89 kg m~(-2) in 2006, with a total soil organic carbon stock(SOCS) decrease of 41.54×10~4 t over the same period. SOCS levels fell most markedly in yellow-brown soil at a rate of51.52%, while an increase of 8.70% was found in the analysed latosol. Geostatistical analysis also showed that the recorded changes in SOCD between 1985 and2006 were spatially structured. The decreasing trend might be attributed to the combined action of intense cultivation, major crop residue removal without any protective tillage measures, unreasonable fertilization and natural climatic diversity inducing a large decrease in SOC in the studied cultivated dry land region of Mojiang County. Therefore, management measures such as protective tillage should be undertaken in order to enhance soil C sequestration.     

SPATIAL AND TEMPORAL DYNAMICS OF SOIL ORGANIC CARBON IN RESERVED DESERTIFICATION AREA —A Case Study in Yulin City, Shaanxi Province, China

Soil organic carbon (SOC) was considered to be a key index in evaluation of soil degradation and soil C sequestration. To discuss the spatial-temporal dynamics of SOC in arable layer in reversed desertification area, a case study was conducted in Yulin City, Shaanxi Province, China. Data of SOC were based on general soil survey in 1982 and repeated soil sampling in 2003. Soil organic carbon content (SOCC) was determined by K2Cr3O2-FeSO4 titration method, and soil organic carbon density (SOCD) was calculated by arithmetic average and area weighted average method, respectively. On average, SOCC and SOCD of the arable layer in the study area from 1982 to 2003 had increased 0.5 Ig/kg and 0.16kg/m2, respectively. Considering main soil types, the widest distributed Arid-Sandic Entisols had lowest values and increments of SOCC and SOCD during the study period; while the second widest Los-Orthic Entisols had higher values and increments of SOCC and SOCD, compared to the mean values of the whole region. The results indicated that reversed desertification process was due to the modification of land use and management practices, such as natural vegetation recovery, planting grass, turning arable land to grassland, and soil and water conservation etc., which can improve SOCC and SOCD and thus enhance soil C sequestration.     

SPATIAL AND TEMPORAL DYNAMICS OF SOIL ORGANIC CARBON IN RESERVED DESERTIFICATION AREA——A Case Study in Yulin City, Shaanxi Province, China

1INTRODUCTIONDesertification is one of the most serious land degrada-tion, which results in the deterioration of physical, che-mical, and biological characteristics of soils (UNEP, 1992). Soil organic carbon (SOC) was considered to be a key index in evaluation of soil quality, soil degradation and soil C sequestration(SCHLENGSINGER etal., 1990; FENG etal., 2002; WANG etal., 2003). Many researchers have reported the correlations among desertification restoration, soil C s…     

Effect of Wetland Reclamation on Soil Organic Carbon Stability in Peat Mire Soil Around Xingkai Lake in Northeast China

Content and density of soil organic carbon(SOC) and labile and stable SOC fractions in peat mire soil in wetland, soybean field and rice paddy field reclaimed from the wetland around Xingkai Lake in Northeast China were studied. Studies were designed to investigate the impact of reclamation of wetland for soybean and rice farming on stability of SOC. After reclamation, SOC content and density in the top 0–30 cm soil layer decreased, and SOC content and density in soybean field were higher than that in paddy field. Content and density of labile SOC fractions also decreased, and density of labile SOC fractions and their ratios with SOC in soybean field were lower than that observed in paddy field. In the 0–30 cm soil layer, densities of labile SOC fractions, namely, dissolved organic carbon(DOC), microbial biomass carbon(MBC), readily oxidized carbon(ROC) and readily mineralized carbon(RMC), in both soybean field and paddy field were all found to be lower than those in wetland by 34.00% and 13.83%, 51.74% and 35.13%, 62.24% and 59.00%, and 64.24% and 17.86%, respectively. After reclamation, SOC density of micro-aggregates( 0.25 mm) as a stable SOC fraction and its ratio with SOC in 0–5, 5–10, 10–20 and 20–30 cm soil layers increased. SOC density of micro-aggregates in the 0–30 cm soil layer in soybean field was 50.83% higher than that in paddy field. Due to reclamation, SOC density and labile SOC fraction density decreased, but after reclamation, most SOC was stored in a more complex and stable form. Soybean farming is more friendly for sustainable SOC residence in the soils than rice farming.     

Influence of climate on soil organic carbon in Chinese paddy soils

Soil organic carbon (SOC) is a major component of the global carbon cycle and has a potentially large impact on the greenhouse effect. Paddy soils are important agricultural soils worldwide, especially in Asia. Thus, a better understanding of the relationship between SOC of paddy soils and climate variables is crucial to a robust understanding of the potential effect of climate change on the global carbon cycle. A soil profile data set (n = 1490) from the Second National Soil Survey of China conducted from 1979 to 1994 was used to explore the relationships of SOC density with mean annual temperature (MAT) and mean annual precipitation (MAP) in six soil regions and eight paddy soil subgroups. Results showed that SOC density of paddy soils was negatively correlated with MAT and positively correlated with MAP (P < 0.01). The relationships of SOC density with MAT and MAP were weak and varied among the six soil regions and eight paddy soil subgroups. A preliminary assessment of the response of SOC in Chinese paddy soils to climate indicated that climate could lead to a 13% SOC loss from paddy soils. Compared to other soil regions, paddy soils in Northern China will potentially more sensitive to climate change over the next several decades. Paddy soils in Middle and Lower Yangtze River Basin could be a potential carbon sink. Reducing the climate impact on paddy soil SOC will mitigate the positive feedback loop between SOC release and global climate change.     

Effects of land cover on soil organic carbon stock in a karst landscape with discontinuous soil distribution

Land cover type is critical for soil organic carbon(SOC) stocks in territorial ecosystems. However, impacts of land cover on SOC stocks in a karst landscape are not fully understood due to discontinuous soil distribution. In this study, considering soil distribution, SOC content and density were investigated along positive successional stages(cropland, plantation, grassland, scrubland, secondary forest, and primary forest) to determine the effects of land cover type on SOC stocks in a subtropical karst area. The proportion of continuous soil on the ground surface under different land cover types ranged between 0.0% and 79.8%. As land cover types changed across the positive successional stages, SOC content in both the 0–20 cm and 20–50 cm soil layers increased significantly. SOC density(SOCD) within 0–100 cm soil depth ranged from 1.45 to 8.72 kg m-2, and increased from secondary forest to primary forest, plantation, grassland, scrubland, and cropland, due to discontinuous soil distribution. Discontinuous soil distribution had a negative effect on SOC stocks, highlighting the necessity for accurate determination of soil distribution in karst areas. Generally, ecological restoration had positive impacts on SOC accumulation in karst areas, but this is a slow process. In the short term, the conversion of croplandto grassland was found to be the most efficient way for SOC sequestration.     

Modified Method for Estimating Organic Carbon Density in Discontinuous Karst Soil Using Ground-Penetrating Radar and Geostatistics

The conventional method which assumes the soil distribution is continuous was unsuitable for estimating soil organic carbon density(SOCD) in Karst areas because of its discontinuous soil distribution. The accurate estimation of SOCD in Karst areas is essential for carbon sequestration assessment in China. In this study, a modified method,which considers the vertical proportion of soil area in the profile when calculating the SOCD, was developed to estimate the SOCD in a typical Karst peak-cluster depression area in southwest China. In the modified method, ground-penetrating radar(GPR) technology was used to detect the distribution and thickness of soil. The accuracy of the method was confirmed through comparison with the data obtained using a validation method, in which the soil thickness was measured by excavation. In comparison with the conventional method and average-soil-depth method,the SOCD estimated using the GPR method showed the minimum relative error with respect to that obtained using the validation method. At a regional scale, the average SOCDs at depths of 0-20 cm and 0-100 cm, which were interpolated by ordinary kriging,were 1.49(ranging from 0.03-5.65) and 2.26(0.09-11.60) kgm-2based on GPR method in our study area(covering 393.6 hm2), respectively. Therefore, the modified method can be applied on the accurate estimation of SOCD in discontinuous soil areas such as Karst regions.     

Storage and density of soil organic carbon in urban topsoil of hilly cities: A case study of Chongqing Municipality of China

Rapid urbanization results in the conversion of natural soil to urban soil,and consequently,the storage and density of the soil carbon pools change.Taking Chongqing Municipality of China as a study case,this investigation attempts to better understand soil carbon pools in hilly cities.First,the vegetated areas in the study area were derived from QuickBird images.Then,topsoil data from 220 soil samples(0-20 cm) in the vegetated areas were collected and their soil organic carbon(SOC) densities were analyzed.Using the Kriging interpolation method,the spatial pattern of SOC was estimated.The results show that the SOC density exhibited high spatial variability in the urban topsoil of Chongqing.First,the SOC density in topsoil decreased according to slope in the order 2°-6° < 25°-90° < 0°-2° < 6°-15° < 15°-25°.Second,the newly developed areas during 2001-2010 had a lower SOC density than the areas built before 1988.Third,urban parks and gardens had a higher SOC density in topsoil,residential green land followed,and scattered street green land ranked last.For hilly cities,the variability of terrain affects the distribution of SOC.The Kriging results indicate that Kriging method combining slope with SOC density produced a high level of accuracy.The Kriging results show that the SOC density to the north of the Jialing River was higher than the south.The vegetated areas were estimated to amount to 73.5 km2 across the study area with an SOC storage of 0.192 Tg and an average density of 2.61 kg/m2.