Abstract：
Spectral indices-based soil prediction models derived from multispectral datasets are too intricate in terms of accuracy as well as resolution. Complications arise while incorporating multispectral datasets for regional-scale spatial assessment of soil macronutrients. Sporadically satellite image fusion techniques have been used for soil nutrient interpolation to circumvent the complications. The fusion of multispectral bands encompasses precise soil information that cannot be observed as accurate with single satellite dataset. In this study, fusion of near infrared regions of Landsat 8 Operational Land Imager and Sentinel 2 has been observed for its contribution on soil macronutrient assessments. Area-to-point regression Kriging (ATPRK) approach is followed in fusing the two satellite imagery and in situ soil spectral have used for the validation of the resultant. Comparative statistical analysis on Landsat 8 OLI band 5 (wavelength: 845-885 nm), Sentine-2 band 8,8A (wavelength: 785-900 nm) datasets and fused satellite bands provides R2 values of 0.8209, 0.8436, and 0.8763 respectively. Regression models y = (0.25006 ± 0.00754) + (0.0000313)x, y = (0.25252 ± 0.0062) + (0.0000810)x, and y = (0.23715 ± 0.0062) + (0.0001210)x for nitrogen, phosphorus, and potassium respectively aids for soil macronutrient interpolation and assessments. Computations reveals the ranges of nitrogen, phosphorus and potassium that floats from 48 to 295 kg/ha, 5.0 to 37 kg/ha, and 32 to 455 kg/ha in the study area. Fusion of satellite imagery by ATPRK approaches in soil macronutrient study at regional scale brings the novelty of the study.
摘要：
从多光谱数据集得出的基于光谱指数的土壤预测模型在准确性和分辨率方面过于复杂。合并多光谱数据集以对土壤常量营养素进行区域尺度的空间评估时会出现并发症。卫星图像融合技术已被用于土壤养分插值，以避免并发症。多光谱波段的融合包含了精确的土壤信息，而这些信息无法用单个卫星数据集观察到。在这项研究中,已观察到Landsat8OperationalLandImager和Sentinel2的近红外区域融合对土壤常量营养素评估的贡献。在融合两个卫星图像和用于验证结果的原位土壤光谱时，遵循了面积到点回归Kriging（ATPRK）方法。Landsat8OLI波段5（波长：845-885nm）的比较统计分析，Sentine-2波段8,8A（波长：785-900nm）数据集和融合卫星波段的R2值分别为0.8209、0.8436和0.8763。回归模型y=(0.25006±0.00754)+(0.0000313)x，y=(0.25252±0.0062)+(0.0000810)x,对于氮，y=(0.23715±0.0062)+(0.0001210)x，磷，和钾分别有助于土壤常量营养素插值和评估。计算揭示了氮的范围，磷和钾从48到295公斤/公顷浮动，5.0至37公斤/公顷，研究区域为32至455公斤/公顷。通过ATPRK方法在区域尺度的土壤常量营养素研究中融合卫星图像带来了研究的新颖性。
