目的 为明确南方丘陵山区水土保持监测中植被覆盖度(FVC)提取的最优遥感数据方案,提升土壤侵蚀模型参数的准确性。 方法 以福建省长汀县为研究区,基于2022年4月获取的6种分辨率遥感影像(2、10、16、30、250、500 m)及9月无人机3 cm验证数据,采用像元二分模型计算NDVI、EVI2、NDMVI、TAVI、SEVI 5类植被指数的FVC,分别从空间分布特征、地形校正效果与精度验证3个方面开展系统对比。 结果 2 m分辨率FVC整体值域偏低且噪声显著;30 m及更低分辨率受混合像元影响,FVC稳定性下降。10 m分辨率在空间细节与噪声控制之间实现最佳平衡。通过无人机多光谱数据验证,10 m NDVI与NDMVI的FVC精度最高,R²均为0.193,MAE均为0.175,优于其他分辨率。NDMVI在多分辨率上表现出稳定的地形校正能力,但10 m NDVI地形校正效果与其视觉差异极小,结合算法简洁性、通用性和处理效率,后者表现出更佳适用性。 结论 10 m Sentinel-2 NDVI兼顾空间细节保留与噪声抑制,是南方丘陵区水土保持业务化监测中FVC提取的优选方案,可为土壤侵蚀模型参数化及区域生态监测提供可靠支撑。
Abstract
Objective This study aims to identify the optimal remote sensing data scheme for extracting fractional vegetation cover (FVC) in soil and water conservation monitoring in southern hilly regions and to improve the accuracy of soil erosion model parameters. Methods Changting County, Fujian Province, was selected as the study area. Based on remote sensing imagery at six resolutions (2, 10, 16, 30, 250, and 500 m) acquired in April 2022 and 3 cm-resolution UAV data collected in September for validation, FVC was calculated for five vegetation indices-NDVI, EVI2, NDMVI, TAVI, and SEVI-using the pixel dichotomy model. A systematic comparison was conducted from three aspects: Spatial distribution characteristics, topographic correction performance, and accuracy validation. Results The FVC at 2 m resolution exhibited an overall lower value range and significant noise, whereas the stability of FVC at 30 m and coarser resolutions decreased due to mixed-pixel effects. The 10 m resolution achieved the best balance between spatial detail preservation and noise control. Validation using UAV multispectral data showed that FVC derived from 10 m NDVI and NDMVI achieved the highest accuracy, with both R² values of 0.193 and MAE of 0.175, outperforming other resolutions. NDMVI exhibited stable topographic correction performance across multiple resolutions; however, the visual difference in the topographic correction performance for the 10 m NDVI was minimal. Considering algorithmic simplicity, general applicability, and processing efficiency, the latter showed better overall suitability. Conclusion The 10 m Sentinel-2 an effective balance between spatial detail retention and noise suppression. It is an optimal option for operational FVC extraction in soil and water conservation monitoring in southern hilly regions, and provides reliable support for soil erosion model parameterization and regional ecological monitoring.
基于此,本研究选取地貌破碎、坡向差异显著且长期作为国家南方红壤水土保持示范区的长汀县作为研究对象,通过系统对比2~500 m 6种分辨率遥感数据,探讨尺度效应对植被指数与FVC计算的影响;并对比NDVI、EVI2、NDMVI、TAVI和SEVI 5种指数在南方丘陵区的地形适应性与稳定性。同时结合无人机3 cm高精度数据开展验证,评估不同指数与不同分辨率的综合表现,优选兼顾空间细节、处理效率与业务化可行性的最优FVC遥感方案。研究成果可为区域土壤侵蚀模型参数化、南方丘陵区水土保持业务化监测及“长汀经验”在更大范围的推广应用提供科学依据。
地形校正效果的对比显示,NDMVI在大多数分辨率下具有较好的地形稳健性,但10 m NDVI在未进行地形校正的情况下,其计算的FVC与NDMVI所计算的FVC差异较小。考虑到算法成熟度高、处理流程简洁、产品稳定性强,同时10 m Sentinel-2 数据在空间分辨率、重访周期及与5 m×20 m径流小区尺度的匹配性方面具有实践优势,进一步增强其在区域监测中的适用性。
从我国土壤侵蚀径流小区的典型尺度看,常用试验地块面积与Sentinel-2的10 m像元尺度高度匹配,有利于在试验尺度与区域尺度之间实现平滑过渡,并兼顾空间分辨率与重访周期的应用需求。结合GEE平台可直接获取预处理数据的便利性,10 m NDVI在精度、稳定性与可操作性上均表现出显著优势,是长汀县丘陵区开展植被覆盖度定量监测的高适配选择。
SARTORIM, FERRARIE, M'BAREKR, et al. Remaining loyal to our soil: A prospective integrated assessment of soil erosion on global food security[J].Ecological Economics,2024,219:e108103.
LANGY, LIUN, LIUS R. Changes in soil erosion and its driving factors under climate change and land use scenarios in Sichuan-Yunnan-Loess Plateau region and the southern hilly mountain belt,China[J].Acta Ecologica Sinica,2021,41(13):5106-5117.
WANGP J, LIUQ, SUNH, et al. Temporal and spatial variation of ecosystem service values in red soil erosion areas in south China[J].Transactions of the Chinese Society for Agricultural Machinery,2021,52(5):219-228.
ZENGL J, YANGZ Q, QINF C, et al. Spatial heterogeneity of vegetation cover change and soil conservation evolution[J].Journal of Soil and Water Conservation,2023,37(5):178-188.
CHENC L, ZHAOG J, MUX M, et al. Spatial-temporal change of soil erosion in Huangshui watershed based on RUSLE model[J].Journal of Soil and Water Conservation,2021,35(4):73-79.
WUS F, ZHANGB, SHIX J, et al. Prediction of soil erosion under different land uses in the typical watershed of the Loess Plateau based on FLUS-CSLE model[J].Transactions of the Chinese Society of Agricultural Engineering,2022,38(24):83-92.
CHENS J, HUANGY Z, TAND G, et al. Prediction and spatial driving force analysis of soil erosion in Badong county based on CA-Markov model[J].Science of Soil and Water Conservation,2025,23(1):222-232.
CHENJ J, HUANGY, ZHAOX N, et al. Accuracy evaluation of vegetation coverage inversion model for alpine grassland in the source region of the Yellow River[J].Science Technology and Engineering,2019,19(15):37-45.
CHIH R, LIY X, WUF, et al. Spectral scale effects on the optical estimation of winter wheat leaf SPAD value[J].Transactions of the Chinese Society of Agricultural Engineering,2025,41(2):196-205.
WUZ J, HEG J, HUANGS L, et al. Terrain effects assessment on remotely sensed fractional vegetation cover in hilly area of southern China[J].Journal of Remote Sensing,2017,21(1):159-167.
ZHANGJ Z, LIC G, WANGT. Dynamic changes of vegetation coverage on the Loess Plateau and its factors[J].Research of Soil and Water Conservation,2022,29(1):224-230.
[24]
ZHAOX Y, TANS C, LIY P, et al. Quantitative analysis of fractional vegetation cover in southern Sichuan urban agglomeration using optimal parameter geographic detector model,China[J].Ecological Indicators,2024,158:e111529.
HOUZ X, JINGC Q, CHENC, et al. Spatiotemporal variation of vegetation coverage of natural grassland in northern Xinjiang in recent 20 years and its relationship with meteorological factors[J].Xinjiang Agricultural Sciences,2023,60(2):464-471.
SHENB B, WEIY B, MAL C, et al. Spatiotemporal changes and drivers of fractional vegetation cover in Inner Mongolia grassland of China[J].Transactions of the Chinese Society of Agricultural Engineering,2022,38(12):118-126.
WANGX L, SHIS H, CHENJ. Change and driving factors of vegetation coverage in the Yellow River basin[J].China Environmental Science,2022,42(11):5358-5368.
YUZ C, JINS L, LIL, et al. The influence of remote sensing data with different spatial resolutions on fractional vegetation cover in soil erosion areas:Case study of Changting County,Fujian Province[J].Subtropical Soil and Water Conservation,2025,37(3):7-10.
[33]
TAOG F, JIAK, WEIX Q, et al. Improving the spatiotemporal fusion accuracy of fractional vegetation cover in agricultural regions by combining vegetation growth models[J].International Journal of Applied Earth Observation and Geoinformation,2021,101:e102362.
PIX Y, ZENGY N, HEC Q. Estimating urban vegetation coverage on the basis of multisource remote sensing data and temporal mixture analysis[J].National Remote Sensing Bulletin,2021,25(6):1216-1226.
[36]
AHMAD ANEESS, ZHANGX L, SHAKEELM, et al. Estimation of fractional vegetation cover dynamics based on satellite remote sensing in Pakistan: A comprehensive study on the FVC and its drivers[J].Journal of King Saud University - Science,2022,34(3):e101848.
CHENM, WANGX Q, LINJ L, et al. Quantitative effects of land use and vegetation cover changes on soil erosion in Changting County in recent 30 years[J].Journal of Soil and Water Conservation,2023,37(5):168-177.
[39]
郭铌.植被指数及其研究进展[J].干旱气象,2003,21(4):71-75.
[40]
GUON. Vegetation index and its advances[J].Journal of Arid Meteorology,2003,21(4):71-75.
[41]
JIANGZ Y, HUETEA, DIDANK, et al. Development of a two-band enhanced vegetation index without a blue band[J].Remote Sensing of Environment,2008,112(10):3833-3845.
WUZ J, XUH Q. A new index for vegetation enhancements of mountainous regions based on satellite image data[J].Journal of Geo-Information Science,2011,13(5):656-664.
JIANGH, HEG J, HUANGH M, et al. Construction and validation of combination model of topography-adjusted vegetation index based on band-ratio model[J].Transactions of the Chinese Society of Agricultural Engineering,2017,33(5):156-161.
JIANGH, YUANY W, WANGS. Shadow-eliminated vegetation index (SEVI) for removing terrain shadow effect:Evaluation and application[J].Journal of Geo-Information Science,2019,21(12):1977-1986.
LIM M, WUB F, YANC Z, et al. Estimation of vegetation fraction in the upper basin of Miyun Reservoir by remote sensing[J].Resources Science,2004,26(4):153-159.
ZHAOA Z, TIANX L. Spatiotemporal evolution and influencing factors of vegetation coverage in the Loess Plateau from 1986 to 2021 based on GEE platform[J].Ecology and Environmental Sciences,2022,31(11):2124-2133.
SONGM L, CHENH T, DINGH, et al. Temporal and spatial variation characteristic and influencing factors of vegetation coverage in Tianjin during 1990—2020[J].Research of Soil and Water Conservation,2023,30(1):154-163.
[54]
WILLMOTTC J, MATSUURAK. Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance[J].Climate Research,2005,30(1):79-82.
[55]
ZHAOY, CHENX Q, SMALLMANT L, et al. Characterizing the error and bias of remotely sensed LAI products:An example for tropical and subtropical evergreen forests in south China[J].Remote Sensing,2020,12(19):e3122.
[56]
CHAIT, DRAXLERR R. Root mean square error (RMSE) or mean absolute error (MAE)?:Arguments against avoiding RMSE in the literature[J].Geoscientific Model Development,2014,7(3):1247-1250.
[57]
LEGATESD R, CABE G JMC. Evaluating the use of "goodness-of-fit" measures in hydrologic and hydroclimatic model validation[J].Water Resources Research,1999,35(1):233-241.
[58]
MARTÍNEZ PRENTICER, VILLOSLADAM, WARDR D, et al. Synergistic use of Sentinel-2 and UAV-derived data for plant fractional cover distribution mapping of coastal meadows with digital elevation models[J].Biogeosciences,2024,21(6):1411-1431.