Research on estimating soil organic matter content in Northeast China based on CARS-IRIV and neural network optimization algorithm
Abstract
Abstract: Choosing appropriate variable screening methods and models can effectively improve the estimation accuracy of soil organic matter content. This article takes the Haicheng Experimental Field of Shenyang Agricultural University as an example to perform SG (Savitzky-Golay) smoothing on the original soil reflectance. It uses iterative preservation of information variables (IRIV), competitive adaptive reweighted sampling (CARS), and CARS-IRIV hybrid methods to reduce dimensionality and extract relevant features from raw spectral data. A hyperspectral inversion model for total organic matter in soil was established using backpropagation neural network (BPNN), sparse optimization (SSA-BPNN), and chaotic sparse optimization BP neural network (CSSA-BPNN). Determine the coefficient (R2) and root mean square error (RMSE) to evaluate the inversion model. The results show that: (1) the optimized model algorithm is better than the unoptimized algorithm; (2) The combination of CARS-IRIV dimensionality reduction method is better than CARS and IRIV dimensionality reduction algorithms in both results and efficiency; (3) The CSSA-BPNN inversion model based on CARS-IRIV dimensionality reduction has the best prediction performance, with a final prediction of soil total organic matter content R2=0.839 and RMSE=1.705. The inversion accuracy is higher than that of SSA-BPNN and BPNN models, which can provide reference for soil nutrient evaluation in the region.
Keywords: UAV hyperspectral; Soil organic matter element inversion; CSSA optimization algorithm; CARS-IRIV dimension reduction; BPNN modeling
DOI: 10.33440/j.ijpaa.20230601.198
Citation: Fang J Y, Xu C Y, Bai J C, Zhu S F, Zhang H G and Yu F H. Research on estimating soil organic matter content in Northeast China based on CARS-IRIV and neural network optimization algorithm. Int J Precis Agric Aviat, 2023; 6(1): 52–60.
Full Text:
PDFReferences
Qiong, H. O. U., et al. Effects of manure substitution for chemical fertilizers on rice yield and soil labile nitrogen in paddy fields of China: A meta-analysis, 2022. Doi: 10.1016/j.pedsph.2022.09.003
Li X N, et al. Soil pollution and site remediation policies in China: A review. Environmental Reviews, 2015, 23.3: 263–274. Doi: 10.1139/er-2014-0073
Chang D, Zhang Y X. Farmland nutrient pollution and its evolutionary relationship with plantation economic development in China. Journal of Environmental Management, 2023, 325: 116589. Doi 10.1016/ j.jenvman.2022.116589
RAYA-MORENO, Irene, et al. Comparing current chemical methods to assess biochar organic carbon in a Mediterranean agricultural soil amended with two different biochars. Science of the Total Environment, 2017, 598: 604–618. Doi: 10.1016/j.scitotenv.2017.03.168
Niu G X, et al. Effects of decadal nitrogen addition on carbon and nitrogen stocks in different organic matter fractions of typical steppe soils. Ecological Indicators, 2022, 144: 109471. Doi: 10.1016/ j.ecolind.2022.109471
LEVI, Nathan; KARNIELI, Arnon; PAZ-KAGAN, Tarin. Airborne imaging spectroscopy for assessing land-use effect on soil quality in drylands. ISPRS Journal of Photogrammetry and Remote Sensing, 2022, 186: 34–54. Doi: 10.1016/j.isprsjprs.2022.01.018
ODEBIRI, Omosalewa, et al. Modelling soil organic carbon stock distribution across different land-uses in South Africa: A remote sensing and deep learning approach. ISPRS Journal of Photogrammetry and Remote Sensing, 2022, 188: 351–362. Doi: 10.1016/j.isprsjprs.2022.04. 026
Hasegawa T, Hasegawa T. Infrared Spectroscopy as a Vibrational Spectroscopy. Quantitative Infrared Spectroscopy for Understanding of a Condensed Matter, 2017: 1–36. Doi: 10.1007/978-4-431-56493-5_1
Qi H J, et al. Evaluating calibration methods for predicting soil available nutrients using hyperspectral VNIR data. Soil and Tillage Research, 2018, 175: 267–275. Doi: 10.1016/j.still.2017.09.006
Huang Y B, et al. Agricultural remote sensing big data: Management and applications. Journal of Integrative Agriculture, 2018, 17(9): 1915–1931. Doi: 10.1016/S2095-3119(17)61859-8
Liu H J, Zhang Y Z, Zhang B. Novel hyperspectral reflectance models for estimating black-soil organic matter in Northeast China. Environmental monitoring and assessment, 2009, 154: 147–154. Doi: 10.1007/s10661-008-0385-4
Yang C B, et al. Study on hyperspectral estimation model of soil organic carbon content in the wheat field under different water treatments. Scientific Reports, 2021, 11.1: 18582. Doi: 10.1038/s41598-021-98143-0
BAUMGARDNER, Marion F., et al. Reflectance properties of soils. Advances in agronomy, 1986, 38: 1–44. Doi: 10.1016/S0065-2113(08)60672-0
Yang C, et al. Study on hyperspectral monitoring model of soil total nitrogen content based on fractional-order derivative. Computers and Electronics in Agriculture, 2022, 201: 107307. Doi: 10.1016/j.compag.2022.107307
TAVARES TR, et al. Spectral data of tropical soils using dry-chemistry techniques (VNIR, XRF, and LIBS): A dataset for soil fertility prediction. Data in Brief, 2022, 41: 108004. Doi: 10.1016/j.dib.2022.108004
DE ALMEIDA MINHONI RT, et al. Multitemporal satellite imagery analysis for soil organic carbon assessment in an agricultural farm in southeastern Brazil. Science of The Total Environment, 2021, 784: 147216. Doi: 10.1016/j.scitotenv.2021.147216
Yun Y, Wang W, Liang Yi. A strategy that iteratively retains informative variables for selecting optimal variable subset in multivariate calibration. Analytica Chimica Acta, 2014, 807: 36–43. Doi: 10.1016/j.aca.2013.11.032
Zhang L, He S, Cheng J, et al. Research on neural network wind speed prediction model based on improved sparrow algorithm optimization. Energy Reports, 2022, 8: 739–747. Doi: 10.1016/j.egyr.2022.09.202
Gao D S, L, Y Z, Xu Q S, et al. A new strategy of outlier detection for QSAR/QSPR. Journal of Computational Chemistry, 2010, 31.3: 592–602. Doi: 10.1002/jcc.21351
Fan S X, Guo Z M, Zhang B H, et al. Using Vis/NIR diffuse transmittance spectroscopy and multivariate analysis to predicate soluble solids content of apple. Food Analytical Methods, 2016, 9.5: 1333–1343. Doi: 10.1007/s12161-015-0313-5
Zhang T T, Zeng S L, Gao Y, et al. Using hyperspectral vegetation indices as a proxy to monitor soil salinity. Ecological Indicators, 2011, 11.6: 1552–1562. Using hyperspectral vegetation indices as a proxy to monitor soil salinity. Ecological Indicators, 2011, 11(6): 1552–1562. Doi: 10.1016/j.ecolind.2011.03.025
He C, Ma M, Wang P. Extract interpretability-accuracy balanced rules from artificial neural networks: A review. Neurocomputing, 2020, 387: 346–358. Doi: 10.1016/j.neucom.2020.01.036
Lu B, Liu N, Li H, et al. Quantitative determination and characteristic wavelength selection of available nitrogen in coco-peat by NIR spectroscopy. Soil and Tillage Research, 2019, 191: 266–274. Doi: 10.1016/j.still.2019.04.015
Wu R, Huang H, Wei J, et al. An improved sparrow search algorithm based on quantum computations and multi-strategy enhancement. Expert Systems with Applications, 2023, 215: 119421. Doi: 10.1016/j.eswa.2022.119421
Zhang C, Ding S. A stochastic configuration network based on chaotic sparrow search algorithm. Knowledge-Based Systems, 2021, 220: 106924. Doi: 10.1016/j.knosys.2021.106924
Gholizadeh A, Saberioon M, Rossel R A V, et al. Spectroscopic measurements and imaging of soil colour for field scale estimation of soil organic carbon. Geoderma, 2020, 357: 113972. Doi: 10.1016/j.geoderma.2019.113972
Deiss L, Margenot A J, Culman S W, et al. Tuning support vector machines regression models improves prediction accuracy of soil properties in MIR spectroscopy. Geoderma, 2020, 365: 114227. Doi: 10.1016/j.geoderma.2020.114227
Refbacks
- There are currently no refbacks.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.