An automated deep learning convolutional neural network algorithm applied for soil salinity distribution mapping in Lake Urmia, Iran

Mohammad Kazemi Garajeh; Farzad Malakyar; Qihao Weng; Bakhtiar Feizizadeh; Thomas Blaschke; Tobia Lakes

doi:10.1016/j.scitotenv.2021.146253

An automated deep learning convolutional neural network algorithm applied for soil salinity distribution mapping in Lake Urmia, Iran

Sci Total Environ. 2021 Jul 15:778:146253. doi: 10.1016/j.scitotenv.2021.146253. Epub 2021 Mar 6.

Authors

Mohammad Kazemi Garajeh¹, Farzad Malakyar², Qihao Weng³, Bakhtiar Feizizadeh⁴, Thomas Blaschke⁵, Tobia Lakes⁶

Affiliations

¹ Department of Remote sensing and GIS, University of Tabriz, Tabriz, Iran. Electronic address: kazemi20.0432@gmail.com.
² Department of Remote sensing and GIS, University of Tabriz, Tabriz, Iran.
³ Center for Urban and Environmental Change, Department of Earth and Environmental Systems, Indiana State University, Terre Haute, IN, USA.
⁴ Department of Remote sensing and GIS, University of Tabriz, Tabriz, Iran; Institute of Environment, University of Tabriz, Iran.
⁵ Geoinformatics Z_GIS, University of Salzburg, Salzburg, Austria.
⁶ Lab of Geoinformation Sciences, Department of Geography, Humboldt University Berlin, Berlin, Germany.

PMID: 33721643
DOI: 10.1016/j.scitotenv.2021.146253

Abstract

Traditional soil salinity studies are time-consuming and expensive, especially over large areas. This study proposed an innovative deep learning convolutional neural network (DL-CNN) data-driven approach for SSD mapping. Multi-spectral remote sensing data encompassing Landsat series images provide the possibility for frequent assessment of SSD in various regions of the world. Therefore, Landsat 7 ETM+ and 8 OLI images were acquired for years 2005, 2010, 2015 and 2019. Totally, 704 sample points collected from the top 20 cm of the soil surface, which 70% was used to train the network and the remains (30%) were utilized to validate the network. Accordingly, DL-CNN model trained using remote sensing (RS)-derived variables (land surface temperature (LST), Soil moisture (SM) and evapotranspiration) and geospatial data such as NDVI and landuse. To train the CNN, ReLu, Cross-entropy and ADAM were employed respectively as activation, loss/cost functions and optimizer. The results indicated the high confidence of OA 0.94.02, 0.93.99, 0.94.87 and 0.95.0 respectively for years 2005, 2010, 2015 and 2019. These accuracies demonstrated the best performance of automated DL-CNN for SSD mapping compared to RS soil salinity indexes. Furthermore, the FR and WOE models applied in order to generate a geospatial assessment of the DL-CNN classification results. According to the FR model, landuse, LST, LST and NDVI with the frequency ratio of 0.98.25, 0.94.03, 0.97.23 and 0.96.36 selected respectively as more effective factors for SSD in the study area for years 2005, 2010, 2015 and 2019. Also based on the WOE model, landuse, LST, landuse and NDVI with the WOE of 0.88.25, 0.91.88, 0.87.43 and 0.89.02 were ranked respectively for years 2005, 2010, 2015 and 2019 as efficient variables for SSD. In sum, our introduced method can be recommended for SDD spatial modelling in other favored areas with similar environmental conditions.

Keywords: Convolutional neural network (CNN); Deep learning (DL); Frequency ratio (FR) model; Lake Urmia; Soil salinity distribution (SSD); Weights of evidence (WOE).