Soil carbon content modeling using machine learning and syntethic soil images from Landsat time series as features. The novelty is the comparison of extensive bare soil indexes proposed for bare soil detection on Landsat images against a fully automatic procedeure using machine learning.
References:
Demattê, J. A. M., Safanelli, J. L., Poppiel, R. R., Rizzo, R., Silvero, N. E. Q., Mendes, W. de S., Bonfatti, B. R., Dotto, A. C., Salazar, D. F. U., Mello, F. A. de O., Paiva, A. F. da S., Souza, A. B., Santos, N. V. dos, Maria Nascimento, C., Mello, D. C. de, Bellinaso, H., Gonzaga Neto, L., Amorim, M. T. A., Resende, M. E. B. de, … Lisboa, C. J. da S. (2020). Bare Earth’s Surface Spectra as a Proxy for Soil Resource Monitoring. Scientific Reports, 10(1), 4461. https://doi.org/10.1038/s41598-020-61408-1
Diek, S., Fornallaz, F., Schaepman, M. E., & Rogier De Jong. (2017). Barest Pixel Composite for Agricultural Areas Using Landsat Time Series. Remote Sensing, 9(12), 1245. https://doi.org/10.3390/rs9121245
Nguyen, C. T., Chidthaisong, A., Kieu Diem, P., & Huo, L.-Z. (2021). A Modified Bare Soil Index to Identify Bare Land Features during Agricultural Fallow-Period in Southeast Asia Using Landsat 8. Land, 10(3), 231. https://doi.org/10.3390/land10030231