-
The study area is located in the coast area of Yancheng City of Jiangsu Province (120.52°N–120.86°N, 33.03°E–33.5°E). This area experiences the subtropical ocean monsoon climate, influenced by continental and marine climates. The annual precipitation is about 900–1000 mm, and the annual temperature is about 13.7°C–14.8°C (Fang et al., 2015). Seawater infiltration leads to groundwater with higher mineralization, and improper utilization of human activities makes shallow saline groundwater rise to the surface, resulting in the increasing soil salinization. Salinization degrades soil quality and makes it difficult to use, so the coastal land is mainly aquaculture land and reclamation area (Werner et al., 2013). Fig. 1 showed the study area and sampling sites.
-
We conducted a field survey for three days on August 2018, and the weather at that time were mostly sunny and a little windy. We drove along the coastline and chose the bare soil surface with different states for photographing and sampling. To increase the complexity of digital image data for developing robust models, we photographed and sampled a total of 52 sites all day long at the sunny and cloudy weather conditions (Fig. 2).
Figure 2. Sampling time for each sampling point in the coast area of Yancheng City of Jiangsu Province, China in 2018
In this field survey, Cannon EOS 760D digital camera with a resolution of 6000 × 4000 pixels was used to take photos. The shooting process is as follows. Firstly, we chose a piece of bare land without the interference of vegetation and other sundry objects. Secondly, height of the camera from the ground was adjusted to keep the bare land area covering the whole field of view (FOV) of the camera. Finally, digital images were captured at automatic mode with the lens perpendicular to the ground. Automatic mode would prevent incorrect camera focusing and exposure time, and help to obtain stable photo quality under various ambient light background in the field (Aitkenhead et al., 2016). Digital images were saved in JPG format. After obtaining soil images, we collected the top soil (about 0–5 cm) and immediately put the soil into a plastic bag. The sampled soils were kept sealed and stored in a dry box until we measured their wet weight in the same day. At the same time the location of sampling points was recorded.
-
After carrying the soil samples into the laboratory, we put them in the oven at 105°C for 24 h to obtain the dry weight. Soil water content is obtained via soil dry weight and wet weight (Jackson, 2005). Soil salt content (SSC) was measured via dry evaporation methods (Rhoades and Ingvalson, 1971). To solve the edge deformation problem of digital images, we cut off the 20% of the image edges, and only consider the central part of the soil images. Digital images in JPG format could be divided into three color components (RGB) and also could be converted to grayscale (Gr). The value range of four color components (R, G, B, Gr) is 0–255. Table 1 gives the summary statistics of soil color components and soil properties.
Table 1. Summary of soil color components and soil properties (n = 52)
Soil property Min. Median Mean Max. SD CV R 59.72 224.12 206.6 250.12 45.18 0.22 G 54.99 210.68 193.74 244.07 44.56 0.23 B 49.03 193.56 178.94 234.35 44.5 0.25 Gr 54.98 210.02 193.09 242.85 44.55 0.23 SSC 0.80 78.87 68.21 131.26 34.62 0.51 SWC 4.06 18.33 18.57 43.88 7.69 0.41 Notes: SD, standard deviation; CV, coefficient of variation; SWC / %, unit of soil water content; SSC / (g/kg), unit of soil salt content, R, G, B and Gr mean red, green, blue, and gray. Hereinafter inclusive -
After we got the soil properties and soil color component information. We calculated the mean of soil image brightness for each sample, and analyzed the correlation coefficient between soil properties and average brightness. Besides, we further analyzed the relationship between soil properties and each brightness level of four color components. The correlation coefficient can be calculated as follow (Wang et al., 2019).
$$ \begin{split} \\ r=\frac{\displaystyle\sum \left(x-\overline{x}\right)\left(y-\overline{y}\right)}{\sqrt{\displaystyle\sum {\left(x-\overline{x}\right)}^{2}\displaystyle\sum {\left(y-\overline{y}\right)}^{2}}} \end{split} $$ (1) where, r means correlation coefficient, x and y are two variables,
$ \overline{x} $ and$ \overline{y} $ are the average value of x and y. -
Each image has at most 256 brightness levels, and each brightness level has a lot of pixels. We counted the pixel number at each brightness level and calculated their proportion to all pixels in an image. For each soil sample, we get soil properties and the pixel proportion of 256 brightness values. However, not all the brightness levels are helpful to model soil properties, so we need to distinguish useful ones from all the brightness levels.
We analyzed the relationship between soil properties and pixel proportion of all brightness values, and found the significantly correlated brightness values. The significance level (P) can be set at P < 0.05, P < 0.01 and P < 0.001. For different significance levels, there will be different correlated brightness levels chosen as the input variables. To test the accuracy of models, we split randomly 70% of dataset as the training data, the rest as the testing data. After standardizing the training data, we applied the random forest algorithm to build the prediction models.
Random forest algorithm is an ensemble learning method that generates many trees (Breiman, 2001). All the trees are trained with the training data by the method of bootstrap sampling, and the accuracy is verified by out-of- bag samples. The algorithm has two main parameters, namely number of decision trees (ntree) and node number (mtry). In this study, the two parameters are set to the default values. We applied 10-fold cross-validation method to verify the accuracy and repeated 10 times to calculated the average value as the final prediction. Then we tested the validation of models with the testing data, and repeated 100 times to identify the optimal models.
In addition, the determining coefficient (R2), root mean square error (RMSE) and the ratio of performance to deviation (RPD) are employed to evaluate the performance of models with both training data and testing data (Xu et al., 2020).
$$ {{{{R}}}}^{2}=1-\frac{\displaystyle\sum _{i=1}^{n}{({P}_{i}-{O}_{i})}^{2}}{\displaystyle\sum _{i=1}^{n}{({P}_{i}-\overline{O})}^{2}} $$ (2) $$ {R}{M}{S} {E}=\sqrt{\frac{1}{n}\sum _{i=1}^{n}{({P}_{i}-{O}_{i})}^{2}} $$ (3) $$ {R}{P} {D}=\sqrt{\frac{\displaystyle\sum _{i=1}^{n}{({O}_{i}-\overline{O})}^{2}}{n\times RMS E}} $$ (4) where O means the observed data, P means the predicted data,
$\overline{O} $ means the mean of observed data, i means the order of each data, n means the number of all data.Generally, models with a higher R2 and a lower RMSE have better predictive power. Models with RPD > 2 have good predictability. Models with 2 > RPD > 1.4 have comparatively general predictability, and models with RPD < 1.4 have less reliable predictability (Chang et al., 2001).
-
Soil salt content was significantly relative with the average image brightness values, and soil water content was insignificantly relative with them (Fig. 3). Also the relationship between soil salt content and soil water content was relatively weak (r = 0.29) but reached a significant level (P < 0.05). The relationship among the average brightness values of four color components was highly relative with each other.
-
For soil salt content, it has highly significant correlation with four color components for many brightness levels, and the best correlation coefficient reached negatively 0.56 at 150 brightness value (P < 0.001) and positively 0.53 at 221 brightness value (P < 0.001). Particularly worth mentioning is that there are generally two brightness ranges closely related with soil salt content, negative correlation at about 110–160 and positive correlation at about 220–250 brightness range (Fig. 4a).
Figure 4. Correlation coefficient (r) between brightness levels and (a) soil salt content and (b) soil water content for each color component in the coast area of Yancheng City of Jiangsu Province, China. the dotted line is r = 0; significance analysis (P) between brightness levels and (c) soil salt content and (d) soil water content for each color component, the dotted line is P = 0.05. R, G, B and Gr mean red, green, blue and gray. SSC and SWC represent soil salt content and soil water content
Different from the average brightness value, soil water content has significantly relationship with some brightness levels for each color component. For the best positive correlations, soil water content was highly related to R, G, B and Gr color components at the brightness value 24 (r = 0.468, P < 0.001), 22 (r = 0.476, P < 0.001), 15 (r = 0.463, P < 0.001), and 24 (r = 0.479, P < 0.001) respectively. As for the best negative correlations, soil water content was closely correlative with R, G, B and Gr color components at the brightness value 161 (r = –0.303, P = 0.029), 150 (r = –0.302, P = 0.03), 127 (r = –0.293, P = 0.035), and 157 (r = –0.304, P = 0.028) respectively. By contrast, the best correlations were mainly concentrated at the 10–30 brightness range (Fig. 4b).
Many brightness levels at about 80–250 were significantly relative with soil salt content, while only a few brightness levels at about 0–40 were significantly relative with soil water content (Figs. 4c and 4d).
-
Based on the above results, we knew that there were many useless brightness levels less relevant to soil properties. Brightness levels selection was hence a necessary step to confirm the relevant variables for modeling. As mentioned earlier, significance analysis provided the correlative brightness levels for four color components, so we determined the relevant variables with P value. In this study, we tried three P values (0.05, 0.01, and 0.001), and calculated the number of brightness levels for four color components at each P value (Table 2). For each P value and color component, one set of data, including soil properties and image brightness data, were generated for modeling and validation. Due to three P values and four color components, there would be twelve (3 × 4 = 12) datasets for modeling each soil property.
Table 2. Related brightness levels for four color components at three significant levels
Soil property P Brightness levels R G B Gr SSC * 131 137 151 136 ** 116 119 129 118 *** 75 76 85 69 SWC * 83 64 54 66 ** 36 34 27 34 *** 25 27 14 26 Notes: R, G, B and Gr mean red, green, blue and gray. SSC and SWC represent soil salt content and soil water content. * means P < 0.05, ** means P < 0.01, *** means P < 0.001 All four color components had enough ability to develop predictable models for soil salt content estimation, and the best one was confirmed with blue color component and at P = 0.05 level. While for soil water content estimation, all four color components could build moderately predictable models with training data, but they were difficult to satisfy the testing data. Under compares, we chose the optimal model with gray color component and at P = 0.01 level for evaluating soil water content (Table 3).
Table 3. Summary of models with each color component at each P value
Soil property Color component P R2m RMSEm / (g/kg) RPDm R2v RMSEv / (g/kg) RPDv SSC R * 0.83 12.20 2.42 0.69 14.77 1.81 ** 0.79 12.45 2.18 0.60 17.26 1.58 *** 0.67 14.02 1.75 0.25 20.99 1.16 G * 0.65 14.09 1.69 0.66 15.18 1.71 ** 0.69 13.95 1.81 0.54 17.22 1.47 *** 0.60 14.34 1.59 0.48 17.85 1.38 B * 0.88 10.64 2.84 0.79 12.00 2.18 ** 0.76 13.86 2.05 0.68 12.55 1.76 *** 0.72 14.44 1.91 0.45 18.65 1.35 Gr * 0.74 14.10 1.97 0.68 15.32 1.77 ** 0.68 14.74 1.78 0.58 15.46 1.53 *** 0.67 15.15 1.75 0.58 16.48 1.54 SWC R * 0.56 3.68 1.51 0.26 4.43 1.16 ** 0.60 3.44 1.59 0.10 4.15 1.05 *** 0.57 3.54 1.52 –0.07 3.79 0.96 G * 0.38 3.53 1.27 –0.68 8.90 0.77 ** 0.62 3.64 1.62 –0.05 4.11 0.98 *** 0.57 3.76 1.53 0.30 2.44 1.20 B * 0.09 4.41 1.05 0.05 3.81 1.03 ** 0.26 4.40 1.17 0.13 2.93 1.07 *** –0.05 4.58 0.98 –0.35 3.94 0.86 Gr * 0.57 3.70 1.52 –0.16 4.02 0.93 ** 0.61 3.60 1.60 0.47 3.04 1.38 *** 0.57 3.67 1.52 0.03 2.79 1.01 Notes: Subscript letter m and v represent the results of modeling and validation. The bold line is the optimal model for estimating SSC and SWC. R, G, B and Gr mean red, green, blue and gray. SSC and SWC represent soil salt content and soil water content. * means P < 0.05, ** means P < 0.01, *** means P < 0.001 Soil salt content inversion model had a higher accuracy with a R2 of 0.79, a RMSE of 12 g/kg, and a RPD of 2.18 for testing data, and the soil water content prediction model had a comparatively lower accuracy with a R2 of 0.47, a RMSE of 3.04 %, and a RPD of 1.38 for testing data (Fig. 5).
Soil Salinity and Soil Water Content Estimation Using Digital Images in Coastal Field: A Case Study in Yancheng City of Jiangsu Province, China
-
Abstract: Soil is the essential part for agricultural and environmental sciences, and soil salinity and soil water content are both the important influence factors for sustainable development of agriculture and ecological environment. Digital camera, as one of the most popular and convenient proximal sensing instruments, has its irreplaceable position for soil properties assessment. In this study, we collected 52 soil samples and photographs at the same time along the coast in Yancheng City of Jiangsu Province. We carefully analyzed the relationship between soil properties and image brightness, and found that soil salt content had higher correlation with average image brightness value than soil water content. From the brightness levels, the high correlation coefficients between soil salt content and brightness levels concentrated on the high brightness values, and the high correlation coefficients between soil water content and brightness levels focused on the low brightness values. Different significance levels (P) determined different brightness levels related to soil properties, hence P value setting can be an optional way to select brightness levels as the input variables for modeling soil properties. Given these information, random forest algorithm was applied to develop soil salt content and soil water content inversion models using randomly 70% of the dataset, and the rest data for testing models. The results showed that soil salt content model had high accuracy (Rv2 = 0.79, RMSEv = 12 g/kg, and RPDv = 2.18), and soil water content inversion model was barely satisfied (Rv2 = 0.47, RMSEv = 3.04%, and RPDv = 1.38). This study proposes a method of modeling soil properties with a digital camera. Combining unmanned aerial vehicle (UAV), it has potential popularization and application value for precise agriculture and land management.
-
Key words:
- soil salinity /
- soil water content /
- coastal soil /
- digital image
-
Figure 4. Correlation coefficient (r) between brightness levels and (a) soil salt content and (b) soil water content for each color component in the coast area of Yancheng City of Jiangsu Province, China. the dotted line is r = 0; significance analysis (P) between brightness levels and (c) soil salt content and (d) soil water content for each color component, the dotted line is P = 0.05. R, G, B and Gr mean red, green, blue and gray. SSC and SWC represent soil salt content and soil water content
Table 1. Summary of soil color components and soil properties (n = 52)
Soil property Min. Median Mean Max. SD CV R 59.72 224.12 206.6 250.12 45.18 0.22 G 54.99 210.68 193.74 244.07 44.56 0.23 B 49.03 193.56 178.94 234.35 44.5 0.25 Gr 54.98 210.02 193.09 242.85 44.55 0.23 SSC 0.80 78.87 68.21 131.26 34.62 0.51 SWC 4.06 18.33 18.57 43.88 7.69 0.41 Notes: SD, standard deviation; CV, coefficient of variation; SWC / %, unit of soil water content; SSC / (g/kg), unit of soil salt content, R, G, B and Gr mean red, green, blue, and gray. Hereinafter inclusive Table 2. Related brightness levels for four color components at three significant levels
Soil property P Brightness levels R G B Gr SSC * 131 137 151 136 ** 116 119 129 118 *** 75 76 85 69 SWC * 83 64 54 66 ** 36 34 27 34 *** 25 27 14 26 Notes: R, G, B and Gr mean red, green, blue and gray. SSC and SWC represent soil salt content and soil water content. * means P < 0.05, ** means P < 0.01, *** means P < 0.001 Table 3. Summary of models with each color component at each P value
Soil property Color component P R2m RMSEm / (g/kg) RPDm R2v RMSEv / (g/kg) RPDv SSC R * 0.83 12.20 2.42 0.69 14.77 1.81 ** 0.79 12.45 2.18 0.60 17.26 1.58 *** 0.67 14.02 1.75 0.25 20.99 1.16 G * 0.65 14.09 1.69 0.66 15.18 1.71 ** 0.69 13.95 1.81 0.54 17.22 1.47 *** 0.60 14.34 1.59 0.48 17.85 1.38 B * 0.88 10.64 2.84 0.79 12.00 2.18 ** 0.76 13.86 2.05 0.68 12.55 1.76 *** 0.72 14.44 1.91 0.45 18.65 1.35 Gr * 0.74 14.10 1.97 0.68 15.32 1.77 ** 0.68 14.74 1.78 0.58 15.46 1.53 *** 0.67 15.15 1.75 0.58 16.48 1.54 SWC R * 0.56 3.68 1.51 0.26 4.43 1.16 ** 0.60 3.44 1.59 0.10 4.15 1.05 *** 0.57 3.54 1.52 –0.07 3.79 0.96 G * 0.38 3.53 1.27 –0.68 8.90 0.77 ** 0.62 3.64 1.62 –0.05 4.11 0.98 *** 0.57 3.76 1.53 0.30 2.44 1.20 B * 0.09 4.41 1.05 0.05 3.81 1.03 ** 0.26 4.40 1.17 0.13 2.93 1.07 *** –0.05 4.58 0.98 –0.35 3.94 0.86 Gr * 0.57 3.70 1.52 –0.16 4.02 0.93 ** 0.61 3.60 1.60 0.47 3.04 1.38 *** 0.57 3.67 1.52 0.03 2.79 1.01 Notes: Subscript letter m and v represent the results of modeling and validation. The bold line is the optimal model for estimating SSC and SWC. R, G, B and Gr mean red, green, blue and gray. SSC and SWC represent soil salt content and soil water content. * means P < 0.05, ** means P < 0.01, *** means P < 0.001 -
[1] Adamchuk V I, Rossel R A V, 2011. Precision agriculture: proximal soil sensing. In: Gliński J et al. (eds.) Encyclopedia of Agrophysics. Encyclopedia of Earth Sciences Series. Dordrecht: Springer, 650–656. doi: 10.1007/978-90-481-3585-1_126 [2] Adamchuk V, Allred B, DoolittleJ et al., 2015. Tools for proximal soil sensing. Soil Survey Staff, C. Ditzler, and L. West, eds. Soil survey manual. United States Department of Agriculture Handbook (18). [3] Adamsen F J, Pinter P J, Barnes E M et al., 1999. Measuring wheat senescence with a digital camera. Crop Science, 39(3): 719–724. doi: 10.2135/cropsci1999.0011183X003900030019x [4] Aitkenhead M, Donnelly D, Coull M et al., 2016. Estimating soil properties with a mobile phone. In: Hartemink A E, Minasny B (eds). Digital Soil Morphometrics, Switzerland: Springer,89–110. doi: 10.1007/978-3-319-28295-4_7 [5] Barrett B W, Dwyer E, Whelan P, 2009. Soil moisture retrieval from active spaceborne microwave observations: an evaluation of current techniques. Remote Sensing, 1(3): 210–242. doi: 10.3390/rs1030210 [6] Breiman L,, 2001. Random forests. Machine Learning, 45(1): 5–32. doi: 10.1023/A:1010933404324 [7] Cantrell K, Erenas M M, De Orbe-Payá I et al., 2010. Use of the hue parameter of the hue, saturation, value color space as a quantitative analytical parameter for bitonal optical sensors. Analytical Chemistry, 82(2): 531–542. doi: 10.1021/ac901753c [8] Chang C W, Laird D A, Mausbach M J et al., 2001. Near-infrared reflectance spectroscopy-principal components regression analyses of soil properties. Soil Science Society of America Journal, 65(2): 480–490. doi: 10.2136/sssaj2001.652480x [9] D’Oleire-Oltmanns S, Marzolff I, Peter K D et al., 2012. Unmanned Aerial Vehicle (UAV) for monitoring soil erosion in Morocco. Remote Sensing, 4(11): 3390–3416. doi: 10.3390/rs4113390 [10] Fang Renjian, Shen Yongming, Shi Haidong, 2015. The changes of coastal wetland landscape pattern based on the characteristics of reclamation: a case study in coastal wetland of Yancheng, Jiangsu province, China. Acta Ecologica Sinica, 35(3): 641–651. (in Chinese) [11] Farifteh J, 2011. Interference of salt and moisture on soil reflectance spectra. International Journal of Remote Sensing, 32(23): 8711–8724. doi: 10.1080/01431161.2010.549522 [12] Fu Y Y, Taneja P, Lin S M et al., 2020. Predicting soil organic matter from cellular phone images under varying soil moisture. Geoderma, 361: 114020. doi: 10.1016/j.geoderma.2019.114020 [13] Islam K, McBratney A, Singh B, 2004. Estimation of soil colour from visible reflectance spectra. Proceedings of Supersoil. [14] Ivushkin K, Bartholomeus H, Bregt A K et al., 2019. Global mapping of soil salinity change. Remote Sensing of Environment, 231: 111260. doi: 10.1016/j.rse.2019.111260 [15] Jackson M L, 2005. Soil chemical analysis: Advanced course. UW-Madison Libraries parallel press. [16] Metternicht G, Zinck J A, 2008. Remote Sensing of Soil Salinization: Impact on Land Management. Boca Raton: CRC Press. [17] Metternicht G I, Zinck J A, 2003. Remote sensing of soil salinity: potentials and constraints. Remote Sensing of Environment, 85(1): 1–20. doi: 10.1016/S0034-4257(02)00188-8 [18] Moonrungsee N, Pencharee S, Jakmunee J, 2015. Colorimetric analyzer based on mobile phone camera for determination of available phosphorus in soil. Talanta, 136: 204–209. doi: 10.1016/j.talanta.2015.01.024 [19] Persson M, 2005. Estimating surface soil moisture from soil color using image analysis. Vadose Zone Journal, 4(4): 1119–1122. doi: 10.2136/vzj2005.0023 [20] Puzachenko Y G, Puzachenko M Y, Kozlov D N et al., 2004. Soil structure analysis with the use of digital color images. Eurasian Soil Science, 37(2): 109–121. [21] Qadir M, Schubert S, Ghafoor A et al., 2001. Amelioration strategies for sodic soils: a review. Land Degradation & Development, 12(4): 357–386. doi: 10.1002/ldr.458 [22] Ren J H, Li X J, Zhao K et al., 2016. Study of an on-line measurement method for the salt parameters of soda-saline soils based on the texture features of cracks. Geoderma, 263: 60–69. doi: 10.1016/j.geoderma.2015.08.039 [23] Rhoades J D, Ingvalson R D, 1971. Determining salinity in field soils with soil resistance measurements. Soil Science Society of America Journal, 35(1): 54–60. doi: 10.2136/sssaj1971.03615995003500010020x [24] Rossel R A V, Behrens T, 2010. Using data mining to model and interpret soil diffuse reflectance spectra. Geoderma, 158(1-2): 46–54. doi: 10.1016/j.geoderma.2009.12.025 [25] Shi T Z, Chen Y Y, Liu H Z et al., 2014. Soil organic carbon content estimation with laboratory-based visible-near-infrared reflectance spectroscopy: feature selection. Applied Spectroscopy, 68(8): 831–837. doi: 10.1366/13-07294 [26] Su C H, Zhang J, Gruber A et al., 2016. Error decomposition of nine passive and active microwave satellite soil moisture data sets over Australia. Remote Sensing of Environment, 182: 128–140. doi: 10.1016/j.rse.2016.05.008 [27] Viscarra Rossel R A, Fouad Y, Walter C, 2008. Using a digital camera to measure soil organic carbon and iron contents. Biosystems Engineering, 100(2): 149–159. doi: 10.1016/j.biosystemseng.2008.02.007 [28] Viscarra Rossel R A, McBratney A B, Minasny B, 2010. Proximal Soil Sensing. Dordrecht: Springer. doi: 10.1007/978-90-481-8859-8 [29] Viscarra Rossel R A, Bouma J, 2016. Soil sensing: a new paradigm for agriculture. Agricultural Systems, 148: 71–74. doi: 10.1016/j.agsy.2016.07.001 [30] Vohland M, Ludwig M, Thiele-Bruhn S et al. , 2014. Determination of soil properties with visible to near- and mid-infrared spectroscopy: effects of spectral variable selection. Geoderma, 223–225: 88–96. doi: 10.1016/j.geoderma.2014.01.013 [31] Wang J Z, Ding J L, Yu D L et al., 2019. Capability of Sentinel-2 MSI data for monitoring and mapping of soil salinity in dry and wet seasons in the Ebinur Lake region, Xinjiang, China. Geoderma, 353: 172–187. doi: 10.1016/j.geoderma.2019.06.040 [32] Wang J Z, Ding J L, Yu D L et al., 2020. Machine learning-based detection of soil salinity in an arid desert region, Northwest China: a comparison between Landsat-8 OLI and Sentinel-2 MSI. Science of the Total Environment, 707: 136092. doi: 10.1016/j.scitotenv.2019.136092 [33] Wang L L, Qu J J, 2009. Satellite remote sensing applications for surface soil moisture monitoring: a review. Frontiers of Earth Science in China, 3(2): 237–247. doi: 10.1007/s11707-009-0023-7 [34] Werner A D, Bakker M, Post V E A et al., 2013. Seawater intrusion processes, investigation and management: Recent advances and future challenges. Advances in Water Resources, 51: 3–26. doi: 10.1016/j.advwatres.2012.03.004 [35] Wu C W, Xia J X, Yang H et al., 2018. Rapid determination of soil organic matter content based on soil colour obtained by a digital camera. International Journal of Remote Sensing, 39(20): 6557–6571. doi: 10.1080/01431161.2018.1460511 [36] Wulf H, Mulder T, Schaepman M E et al. , 2014. Remote Sensing of Soils. Switzerland: University of Zurich. [37] Xu C, Zeng W Z, Huang J S et al., 2016. Prediction of soil moisture content and soil salt concentration from hyperspectral laboratory and field data. Remote Sensing, 8(1): 42. doi: 10.3390/rs8010042 [38] Xu L, Zheng C L, Wang Z C et al., 2019. A digital camera as an alternative tool for estimating soil salinity and soil surface roughness. Geoderma, 341: 68–75. doi: 10.1016/j.geoderma.2019.01.028 [39] Xu L, Viscarra Rossel R A, Lee J et al., 2020. A simple approach to estimate coastal soil salinity using digital camera images. Soil Research, 58(8): 737–747. doi: 10.1071/SR20009 [40] Xu L, Wang H, Qiu S Y et al., 2021a. Coastal soil salinity estimation based digital images and color space conversion. Spectroscopy and Spectral Analysis, 41(8): 2409–2414. (in Chinese) [41] Xu L, Wang Z C, Hu J S et al., 2021b. Estimation of soil salinity under various soil moisture conditions using laboratory based thermal infrared spectra. Journal of the Indian Society of Remote Sensing, 49(4): 959–969. doi: 10.1007/s12524-020-01271-9 [42] Yang R M, Guo W W, 2019. Using Sentinel-1 imagery for soil salinity prediction under the condition of coastal restoration. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 12(5): 1482–1488. doi: 10.1109/JSTARS.2019.2906064 [43] Yashchenko A S, Bobrov P P, 2016. Impact of the soil moisture distribution in the top layer on the accuracy moisture retrieval by microwave radiometer data. IEEE Transactions on Geoscience and Remote Sensing, 54(9): 5239–5246. doi: 10.1109/TGRS.2016.2559162 [44] Yue J B, Tian J, Tian Q J et al., 2019. Development of soil moisture indices from differences in water absorption between shortwave-infrared bands. ISPRS Journal of Photogrammetry and Remote Sensing, 154: 216–230. doi: 10.1016/j.isprsjprs.2019.06.012 [45] Zanetti S S, Cecílio R A, Alves E G et al., 2015. Estimation of the moisture content of tropical soils using colour images and artificial neural networks. Catena, 135: 100–106. doi: 10.1016/j.catena.2015.07.015 [46] Zhu Y J, Wang Y Q, Shao M A et al., 2011. Estimating soil water content from surface digital image gray level measurements under visible spectrum. Canadian Journal of Soil Science, 91(1): 69–76. doi: 10.4141/cjss10054