4.2 8.

Communication

Advanced Soil Organic Matter

Prediction with a Regional Soil NIR
Spectral Library Using Long Short-
Term Memory–Convolutional
Neural Networks: A Case Study

Tianyu Miao, Wenjun Ji, Baoguo Li, Xicun Zhu, Jianxin Yin, Jiajie Yang, Yuanfang Huang, Yan Cao,
Dongheng Yao and Xiangbin Kong

https://doi.org/10.3390/rs16071256
 remote sensing
Communication
Advanced Soil Organic Matter Prediction with a Regional Soil
NIR Spectral Library Using Long Short-Term
Memory–Convolutional Neural Networks: A Case Study
Tianyu Miao 1 , Wenjun Ji 1, * , Baoguo Li 1 , Xicun Zhu 2 , Jianxin Yin 1 , Jiajie Yang 3 , Yuanfang Huang 1 ,
Yan Cao 1 , Dongheng Yao 1 and Xiangbin Kong 1

1 College of Land Science and Technology, China Agricultural University, Beijing 100193, China;
[email protected] (T.M.); [email protected] (B.L.); [email protected] (J.Y.);
[email protected] (Y.H.); [email protected] (Y.C.); [email protected] (D.Y.);
[email protected] (X.K.)
2 College Resources and Environment, Shandong Agricultural University, Taian 271001, China;
[email protected]
3 Institute of Botany, Chinese Academy of Sciences, Beijing 100093, China; [email protected]
* Correspondence: [email protected]

Abstract: Soil analysis using near-infrared spectroscopy has shown great potential to be an alternative
to traditional laboratory analysis, and there is continuously increasing interest in building large-scale
soil spectral libraries (SSLs). However, due to issues such as high non-linearity in soil spectral
data and complexity in soil spatial variation, the establishment of robust prediction models for soil
spectral libraries remains a challenge. This study aimed to investigate the performance of deep
learning algorithms, including long short-term memory (LSTM) and LSTM–convolutional neural
networks (LSTM–CNN) integrated models, to predict the soil organic matter (SOM) of a provincial-
scale SSL, and compare it to the normally used local weighted regression (LWR) model. The Hebei
soil spectral library (HSSL) contains 425 topsoil samples (0–20 cm), of which every 3 soil samples
Citation: Miao, T.; Ji, W.; Li, B.; Zhu, were collected from dry land, irrigated land, and paddy fields, respectively, in different counties
X.; Yin, J.; Yang, J.; Huang, Y.; Cao, Y.; of Hebei Province, China. The results show that the accuracy of the validation dataset rank as
Yao, D.; Kong, X. Advanced Soil follows: LSTM–CNN (R2 p = 0.96, RMSEp = 1.66 g/kg) > LSTM (R2 p = 0.83, RMSEp = 3.42 g/kg) >
Organic Matter Prediction with a LWR (R2 p = 0.82, RMSEp = 3.79 g/kg). The LSTM–CNN model performed the best, mainly due
Regional Soil NIR Spectral Library to its comprehensive ability to effectively extract spatial and temporal features. Meanwhile, the
Using Long Short-Term LSTM model achieved higher accuracy than the LWR model, owing to its built-in memory unit
Memory–Convolutional Neural
and its advantage of faster feature band extraction. Thus, it was suggested to use deep learning
Networks: A Case Study. Remote Sens.
algorithms for SOM predictions in SSLs. However, their performance on larger-scale SSLs such as
2024, 16, 1256. https://doi.org/
continental/global SSLs still needs to be further investigated.
10.3390/rs16071256

Academic Editor: Jeroen Meersmans Keywords: deep learning; long short-term memory (LSTM); long short-term memory–convolutional
neural networks (LSTM–CNN); near-infrared (NIR); soil spectral library
Received: 20 February 2024
Revised: 28 March 2024
Accepted: 28 March 2024
Published: 2 April 2024
1. Introduction
Soils are vital elements of the ecosystem, and have a crucial part to play in both human
survival and environmental sustainability [1,2]. With the explosive growth of the world’s
Copyright: © 2024 by the authors. population and rapid urbanization, the human demand for soil to provide food, fiber, and
Licensee MDPI, Basel, Switzerland.
other resources is increasing; thus, soil resources are being intensively used, leading to
This article is an open access article
various security concerns such as soil degradation and salinization [3–5]. These problems
distributed under the terms and
lead to abnormal levels of organic matter in soils, which pose a threat to agro-ecological
conditions of the Creative Commons
balances, food security, and furthermore impact achievement of the UN Sustainable Devel-
Attribution (CC BY) license (https://
opment Goals (SDGs) [6]. To ensure the safety of soil organic matter and guide the rational
creativecommons.org/licenses/by/
use of soil, the demand for rapid, dynamic soil monitoring is rapidly increasing [7]. In
4.0/).

Remote Sens. 2024, 16, 1256. https://doi.org/10.3390/rs16071256 https://www.mdpi.com/journal/remotesensing

Remote Sens. 2024, 16, 1256 2 of 13

contrast to traditional laboratory analyses, which are not only costly but also environmen-
tally burdensome, soil spectroscopy presents a less resource-intensive alternative. This
technique enables the analysis of various soil property categories from a single sample,
facilitating material savings and promoting sustainable soil management [8]. When collect-
ing soil spectral data, implementing control methods can significantly reduce errors and
enhance the quality of the collected spectra. These methods include (i) standardizing the
size of soil particles to ensure consistency, (ii) repeating the spectrum collection process
multiple times to verify data reliability, and (iii) utilizing machine learning algorithms to
select representative soil samples. Together, these strategies contribute to the precision
and accuracy of soil spectral analysis [9,10]. Previous studies have shown that the use of
near-infrared spectroscopy (NIR, 780–2500 nm) can accurately characterize the component
information of soils [11–16]. The overtones and combinations of fundamental vibrations
are the basis of using NIR spectra for predicting soil properties. As energy quanta are
directly related to frequency (and inversely related to wavelength), the resulting absorption
spectrum has a characteristic shape that can be used for analytical purposes.
Soil spectral libraries (SSLs), established at various geographical scales, play a crucial
role in monitoring soil properties. These libraries, ranging from global to continental,
as well as national and regional scales, significantly ease the process of acquiring soil
information. Their availability facilitates enhanced real-time and dynamic monitoring of
soil safety [17–26]. However, due to the heterogeneity of soil types, climate, topography,
and vegetation within each region, it is difficult to extrapolate from predictive models
developed within a given region to other study areas. This is mainly due to the redundancy
of soil sample datasets, resulting in increased complexity in soil spatial variation and
non-linearity of soil properties and spectral data [15]. With the development of data mining
techniques in recent years, there are many studies applying various linear and non-linear
machine learning algorithms to optimize the prediction of soil spectral library attributes.
Ramirez-Lopez et al. (2013) used the Global Soil Visible-Near-infrared Spectrum Library to
test different distance measurement algorithms [27]. Shi et al. (2014) used soil samples from
the Chinese National Spectral Library for soil spectral classification. They used the linear
model PLSR algorithm to predict the soil organic matter (SOM), resulting in improved
accuracy (R2 = 0.89; RPD = 3.15) [23]. Rossel et al. (2016) developed and analyzed a
global library of visible near-infrared (vis-NIR) spectra for soils, and the conclusions of the
experiment showed that using machine learning algorithms to model the different spectra
allowed finding local relationships in the data to accurately predict soil properties [28].
Zhong et al. (2021) explored the modeling potential of deep convolutional neural networks
(DCNNs) for soil properties based on a large SSL [29]. Wang et al. (2022) evaluated four
spectral pre-processing models and seven machine learning methods using a continental-
scale SSL for model evaluation, and found that the SOM models of mineral and organic soils
should be distinguished on the basis of their distinct spectral signatures [30]. Especially in
larger SSLs, such as global- and national-scale SSLs, most experiments have been applied
to linear models and convolutional neural network models. There are few studies on short-
term memory neural networks, and there is a lack of studies comparing soil predictions
using neural network deep learning algorithms and local weighted regression algorithms.
And there are also fewer studies on provincial-scale SSLs, which could better balance the
relationship between the number of samples and data accuracy of large-scale SSLs [31–37].
To address the above scientific research questions, this study aimed to (i) investigate
the performance of deep learning algorithms, including long short-term memory (LSTM)
and long short-term memory–convolutional neural networks (LSTM–CNN) integrated
models to predict the soil organic matter (SOM) of a provincial-scale SSL, and (ii) compared
it to the normally used local weighted regression (LWR) model.
Remote Sens. 2024, 16, 1256 3 of 13

2. Materials and Methods

2.1. The Hebei Soil Spectral Library
Hebei Province is located in northern China and covers a total area of 188,000 km2 ,
with a gradual rise in terrain from the southeast to the northwest, and an average altitude
of 1200 to 1500 m above sea level. The region has a temperate continental monsoon
climate with an average annual temperature of 10–20 ◦ C, and an average annual rainfall of
484.5 mm. According to the World Soil Classification System (WRB), the main soil types in
Hebei province are anthrosol, cambisol, calcaric, and fluvisol soils [38].
Figure 1 shows the locations of the sampling points. The Hebei Soil Spectral Library
(HSSL) contains 440 samples. The samples were collected from highlands and plains, and
ff
contain three different secondary land use classifications including paddy fields, dry land,
and irrigated land. Using the method of uniform sampling, the soil types, landforms, and
land use types at various sampling points were selected across each county, ensuring a
representative mix. In this process, three soil samples were collected from each county.
The soil samples were collected to a depth of 20 cm. Prior to laboratory analysis and
spectroscopic measurements, all of the soil samples were air dried and ground, passed
through a 2 mm diameter sieve, and split into two using the quadrat method, one part for
soil property determination and the other part for soil spectral data collection. The SOM
was determined with the H2 SO4 –K2 Cr2 O7 oxidation method at 180 ◦ C Celsius for 5 min,
and SOM data were recorded for 431 soil samples [39].

Figure 1. The location of the study area and soil sampling sites.

2.2. Spectra Measurement

Soil NIR spectra were measured using a Bruker NIR spectrometer in a dark room. The
spectral range was 12,800–4000 cm−1− (780 nm–2500 nm) with a resolution of 2 cm−−1 . To
improve the signal-to-noise ratio and minimize interference, each soil sample underwent
three random measurements. These measurements were averaged to yield a representative
spectrum, thereby enhancing the accuracy and reliability of the soil analysis.

2.3. Spectral Pre-Processing

The spectral edge noise was first removed, leaving the 800–2500 nm region. Given
that spectral outliers can impede subsequent data analysis, their influence on the model
was minimized by calculating the Mahalanobis distance (MD) to identify and remove these
outliers [40]. Figure 2 shows the use of the Mahalanobis distance to remove outliers; a total
of 6 abnormal samples were removed.
Remote Sens. 2024, 16, 1256 4 of 13

Figure 2. Outliers removal based on Mahalanobis distance.

In addition, to amplify information and further reduce noise, the following meth-
ods were applied to spectral pretreatment: (i)tzSavitzky–Golay smoothing: A window of
size 5 and a polynomial of order 2 were chosen for smoothing to achieve maximum noise
ff
removal [41]. (ii) First derivatives: First derivatives were used to achieve the effect of en-
hancing the detailed features of a spectrum and eliminating baseline drift and background
noise interference [42].
After spectral pre-processing, the data were modeled to verify the performance of
all models.

2.4. Predictive Algorithms

2.4.1. Locally Weighted Regression
The LWR model is a non-parametric regression method based on locally weighted
regression, proposed by Cleveland and later developed by Næs et al. (1988, 1992) [43,44].
The model estimates the local nature of the data using a weighted least squares regression,
which allows the data to be modeled on otherwise different subsets of the data. The
sample points that are closer to the prediction point haveff a higher weight attached to them,
thus better reflecting the similarity between the sample points. The ideattin soil attribute
predictiontt is to find the closest sample to the predicted sample in the spectral library,
tt and
to fit a local linear regression equation with that sample.
To obtain the best model performance, cross-validation was used for parameter se-
lection. The root mean square error (RMSE) of the model was calculated using 5-fold
cross-validation, with different numbers of K and PLS components as an evaluation metric.
ff
Finally, the corresponding K and PLS component numbers that minimized the RMSE were
selected as the optimal parameters. In this way, the optimal parameters for modeling were
obtained to achieve the best prediction results.

2.4.2. Long Short-Term Memory (LSTM) Model

LSTM is a deep learning model for sequential data analysis, and is considered to
be an extension of recurrent neural networks (RNN), proposed by Sepp Hochreiter and
Jürgen Schmidhuber in 1997 [45]. LSTM can effectively solve the problem of gradient
disappearance and gradient explosion during ff reverse transfer in ordinary convolutional
neural networks [46].
This study used an LSTM model for SOM prediction. The model consists of an input
layer, a hidden layer, and an output layer whose model memory is dependent on one
or more memory units of the hidden layer. A memory cell (Figure 3) can be regarded
as a neuron with the ability to remember; each cell contains an internal state c and an
external state h, representing the long-term and short-term memory of the𝑐cell, respectively;
the memory
ℎ cell state ct indicates that the model retains information accumulated from
previous time steps; the hidden state ht indicates the output at the current time step.
 o
𝑡
Remote Sens. 2024, 16, 1256 ℎ 5 of 13

Figure 3. Architecture of a typical LSTM model. f ,𝑓i, and

𝑖 o denote o forgettingttgates, input gates, and
𝑥
output gates, respectively; xt denotes input at time step t ; h t −1 𝑡 t ℎ
, h ℎ output; ct−1 and c𝑐t denote
denotes
𝑐 states.
cell

The memory cell can decide which information to retain at the previous time step
(t − 1) and what information 𝑓 =to𝜎 output
𝑊 𝑥 +at𝑊theℎ current
+ 𝑊 time
𝑐 + 𝑏 (t), which is controlled
step
mainly by three gates: the forgetting gate ( f ), the input gate (i), and the output gate (o). The
forgetting gate ( f ) (Equation = 𝜎determines
𝑖 (1)) 𝑊 𝑥 + 𝑊 whichℎ +information
𝑊 𝑐 + 𝑏should be discarded from
the memory cell before entering the current cell state (c); the input gate (i) (Equation (2))
controls which information from 𝑐̃ =the
𝑡𝑎𝑛ℎ 𝑊 𝑥 cell
memory +𝑊 ℎ is updated
state +𝑏 during the current time
step, which is calculated by a weighted average of the input data and the previous cell
state (Equations (3) and (4)); finally,𝑐 = ⊗ 𝑐 gate
the𝑓 output +𝑖 ⊗ (ot𝑐̃) (Equation (5)) determines how
much information in the memory cell should be output from the hidden state at the current
moment (t); the final hidden𝑜 state
= 𝜎 (h𝑊t ) (Equation
𝑥 + 𝑊 ℎ (6))+is𝑊 the𝑐result
+ 𝑏 of the information in the
memory cell being weighted by the output gate.
ℎ = 𝑜 ⊗ 𝑡𝑎𝑛ℎ 𝑐
f t = σ (W f i x t + W f h h t − 1 + W f c c t − 1 + b f ) (1)
𝑊 𝑏 ffi
𝜎 it = σ(Wix xt + Wih ht−1 + Wictanℎ c t − 1 + bi ) (2)
⊗
c̃t = tanh(Wcx xt + Wch ht−1 + bc ) (3)
ct = f t ⊗ ct−1 + it ⊗ c̃t (4)
ot = σ(Wox xt + Woh ht−1 + Woc ct + bo ) (5)
ht = ot ⊗ tanh(ct ) (6)
where W and b are the corresponding weight coefficient matrices and bias terms, respec-
tively; σ denotes the logistic sigmoid function; tan h is a hyperbolic tangent activation
tt
function; and the multiplication of elements between vectors is denoted as ⊗.
In the SOM prediction process, the time step can be interpreted as the step length of
the spectral band, rather than the time interval in the traditional time series. Specifically,
each spectral band is treated as a distinct time step within the series.
To enable the model to predict the target value more accurately, the Adam was selected
as the optimizer to update the model parameters within the model [47]. This is an adaptive
learning rate optimization algorithm that adaptively adjusts the learning rate according to
the gradient of each parameter, and has a faster convergence rate compared to the traditional
stochastic gradient descent (SGD); moreover, it enables better model generalization. In
addition to the optimizer, the mean square error (MSE) was chosen as the loss function
to measure the error of the model, allowing calculation of the mean quadratic deviation
between the predicted and true values of the model. By minimizing the MSE loss function,
the model can be made to become a more accurate predictor of target values. Finally, the
rectified linear unit (RELU) was used in the models as an activation function to improve
the non-linear properties of the model [48]. The non-linear transformation helps the model
learn the spectral multicollinearity, thus solving the spectral multicollinearity existing
Remote Sens. 2024, 16, 1256 6 of 13

between the soil samples and the spectral data, andtt adding a dropout layer to prevent the
model from overfitting [49]. For model hyperparameter selection, according to the theory
of Dominic Masters and Carlo Luschi et al. (2018) [50], a batch size of 2–32 can provide
more up-to-date gradient calculations, which can lead to more stable and reliable training
and enable the model to obtain the best performance.

2.4.3. LSTM–CNN Integration Model

Figure 4 displays the LSTM–CNN integration model that was created. The basic
concept is to use the CNN model to extract the background features of the embedded space,
and the LSTM model to extract the temporal features of the variables. After analyzing
a soil spectrum, the relevant bands are immediately extracted, the extracted spatial and
temporal features are linked, and the output is then computed using the fully connected
layer (e.g., the expected SOM).
This integrated model encompasses complex iterative calibration steps, wherein the pa-
rameters and methodologies deployed by LSTM are akin to those described in Section 2.4.2
of this study. The CNN captures spatial features of the data using convolution and pooling
tt
layers. Convolution layers can detect local patterns in the data, while pooling layers can
reduce the dimensionality of the data and preserve important features. Within this integra-
tion model, the outputs from the LSTM and CNN are merged via concatenation operations,
ensuring the preservation of both the temporal dependencies identified by the LSTM and
the spatial features ascertained by the CNN.

Figure 4. Flowchart of LSTM–CNN model iteration.

The LSTM–CNN model was built in Python 3.7 using Keras with Tensorflow as the
backend. The other two models were also built using Python.

2.5. Model Evaluation

In this study, the data set was divided into two parts: the training set and the validation
set. The soil samples were ranked from lowest to highest in terms of SOM content, with the
training and validation sets representing 80% and 20% of the total data set, respectively. In
the model evaluation, the coefficient of determination (R2 ) and the root mean square error
(RMSE) were used as the main indicators for evaluation, where the better the performance
of the model, the higher the R2 and the lower the RMSE.
Remote Sens. 2024, 16, 1256 ffi 7 of 13
tt

 – 2
∑ ∑ŶaŶ −−YY‾p
R2 R= 1=−1 −  (7)
∑ Y −–Y‾ 2
∑ Ya − Ya
s ∑ y − ŷ
RMSE = ∑ni=1 (y − ŷ )2
an a
RMSE = (8)
n
n ‾
– Y – Y ‾
where n is the total numberŶ of samples,
Y Yp and Y a are the average predicted and measured
values, respectively; Ŷa and Ya are the predicted and measured values, respectively.

3. Results
3.1. Descriptive Statistical Results
Table 1 shows the SOM statistics recorded in the HSSL. The soil samples in the HSSL
have a wide range of SOM, distributed between 2.76 and 58.30 g/kg, with an average of
ffi
16.69 g/kg, showing strong heterogeneity with a coefficient of variation (CV) of 51.34%.
The skewness is 2.03.

Table 1. Soil property statistics for the HSSL.

Soil Property No. Min Max Mean Std Skewness CV *

SOM (g/kg) 425 2.76 58.30 16.69 8.57 2.03 51.34
ffi
* CV: coefficient of variation.

ff levels of organic
Figure 5 displays the absorbance spectra of soil samples with different
tt
matter content in the HSSL. From the representative samples selected on the graph, it can
be clearly observed that the value of the absorbance curve is higher for samples with a
high SOM; when the SOM is low, the absorbance is lower. The results indicate a positive
correlation between absorbance and SOM, with absorbance increasing as the SOM increases.
Clear water absorption bands were observed at around 1400 and 1900 nm; these bands are
formed due to the hydroxyl groups in clay minerals present in the soil.

Figure 5. Soil spectra with different SOM content levels in the Hebei soil spectral library.

3.2. Models for SOM Content Prediction of HSSL

The LWR model determined the optimal parameters through cross validation. Figure 6
shows the scatter plot of the model prediction results for the HSSL (validation R2 p = 0.82,
RMSEp = 3.79 g/kg).
 ff
ff

Remote Sens. 2024, 16, 1256 tt

tt 8 of 13

tt
tt plot of SOM prediction results of the validation set using the LWR algorithm.
Figure 6. Scatter
tt
tt plot of the LSTM model used to predict SOM in the HSSL.
Figure 7 shows the scatter
The model used a batch size of 8, two LSTM layers, and a dropout rate of 0.1. This configu-
tt
tt training and convergence of the model, allowing the inversion of
ration facilitated better
soil physical and chemical properties. The results indicate that the R2 values for both the
training and validation sets are greater than 0.8. The training set has an R2 value of 0.86,
while the validation set has an R2 value of 0.83 (as shown in Figure 7). Furthermore, the
validation set has an RMSE of 3.42 g/kg.

(a) (b)
tt
Figure 7. Scatter plots of SOM prediction results of the validation (a) and training (b) sets using the
tt
LSTM algorithm.

The importance index of each band in the range of 800–2500 nm was extracted using
the updated LSTM model selected by memory cells (Figure 8). After ranking, it was found
that the most obvious responses to SOM were in the 800–900 nm, 1800–1900 nm, and
2200–2400 nm bands [14,51,52].
Figure 9 shows the prediction results of the LSTM–CNN integrated model for SOM
in the HSSL. The training results were significantly improved compared to the previous
two models, with both the training and validation sets achieving accuracies of over 0.95.
The R2 of the training set is 0.99, and the R2 of the validation set is 0.96. In addition, the
RMSE of the training set is 0.47 g/kg, and the RMSE of the validation set is 1.66 g/kg. The
integrated model produced more accurate results than other deep learning and machine
learning models used by Wang et al. (2022) [27]. Additionally, the LSTM–CNN model
outperformed other models in similar soil spectral libraries studies [17,36].
Remote Sens. 2024, 16, 1256 9 of 13

Figure 8. Important features of the soil spectrum extracted by the LSTM model.

(a) (b)
tt
Figure 9. Scatter plots of SOM prediction results of the validation (a) and training (b) sets using the
LSTM–CNN algorithm.

Table 2 shows the accuracy of the three models. The R2 p order among the three models
is LSTM–CNN > LSTM > LWR, and the RMSEp order is LWR > LSTM > LSTM–CNN. From
these two indexes, it can be seen that the convolutional neural network model is superior to
the locally weighted regression model overall, and the LSTM–CNN integrated model has
the best effectffamong the three models. In comparison, the LSTM–CNN integrated model
demonstrates enhanced generalizability, evidenced by its superior predictive performance
on soil samples
tt
with high coefficients offfivariation in the spectral data from Hebei province.
This model can be better applied in tt provincial-scale spectral libraries.

Table 2. Table of SOM accuracy for the three prediction models.

Model R2 c RMSEc R2 p RMSEp

LWR - - 0.82 3.79 g/kg
LSTM 0.86 2.81 g/kg 0.83 3.42 g/kg
ff
LSTM–CNN 0.99 0.47 g/kg 0.96 1.66 g/kg

ffi
4. Discussion
tt
4.1. The Application Ability of Provincial-Scale Soil Spectral Library
This study used 425 representative soil samples in the HSSL, covering 13 counties
in Hebei Province. The performances of the three models—LSTM–CNN, LSTM, and
LWR—were evaluated, demonstrating their efficacy in predicting soil organic matter (SOM).
All of the models achieved validation R2 values above 0.8. Specifically, the LSTM–CNN
model yielded the highest accuracy (R2 p = 0.96, RMSE = 1.66 g/kg), followed by the LSTM
(R2 p = 0.83, RMSE = 3.42 g/kg), and LWR (R2 p = 0.82, RMSE = 3.79 g/kg) models. These
results underscore the potential of spectral databases for rapid, non-destructive soil analysis
Remote Sens. 2024, 16, 1256 10 of 13

within the study area, paving the way for real-time dynamic soil property monitoring across
Hebei Province. This study also highlights the balance between sample quantity and data
accuracy afforded by a spectral database of this scale, contributing valuable insights into
soil characteristics within the region.
Based on the results of this study, it was found that the SOM from the HSSL can be
successfully predicted using a deep learning model or machine learning model, which is in
line with previous research [53,54]. The prediction of SOM in the NIR region is a frequently
evaluated metric, mainly because the overtone and combination bands of organic molecules
occur in this band. The bands around 1100 nm, 1600 nm, 1700 nm~1800 nm, 2000 nm, and
2200 nm~2400 nm are recognized as the most important bands for SOM prediction, which
correspond to the feature bands extracted by the LSTM model in this study, which can
confirm the good predictive ability of the LSTM model for soil spectra.
The accuracy of all three algorithms surpassing 0.8 demonstrates the successful appli-
cation of a provincial-scale spectral library, enabling the comprehensive and dynamic mon-
itoring of soil information throughout the province. This approach effectively addresses
the limitations of field-scale spectral libraries and laboratory-based soil property analysis,
offering enhanced detail and representativeness of soil information at a larger scale.

4.2. The Potential of Deep Learning Models for Soil Spectral Property Prediction
Among the evaluated models, the LSTM–CNN model demonstrated superior fitting
capabilities, particularly in managing the connections between complex variables. This
model leverages a unique memory unit to understand long-term dependencies within
sequences, enabling effective screening and modification of sequence issues by integrating
both long-term and short-term memory. Compared to traditional linear models, the LSTM–
CNN model exhibits enhanced stability in extracting features from complex soil spectral
data, thereby simplifying the extraction of feature bands [55,56].
Furthermore, the spectral library’s larger area range often presents nonlinear rela-
tionships due to spectral multicollinearity and data redundancy. This study incorporated
the ReLu algorithm into both the LSTM and LSTM–CNN models, facilitating a non-linear
transformation that addresses the challenge of high non-linearity in spectral data analysis.
As a result of its great ability to fully extract spatial and temporal characteristics, the
LSTM–CNN model is better able to complete relevant soil prediction over a wide variety
of soil spectral libraries. The LSTM–CNN model created in this research is a promising
technology for soil prediction, as few similar studies in the past have used it to perform
SOM prediction in provincial-scale soil spectral libraries.
This model’s hyperparameters were tuned in this study. However, additional research
is required to explore the optimal hyperparameter selection algorithm for both the LSTM
and LSTM–CNN models. This will improve the models’ performance and broaden their
application in predicting soil spectral properties. Further research could enhance our
understanding of the performance of these models and better support their application in
predicting soil spectral properties.

5. Conclusions
In previous studies related to large-scale SSLs, the majority of methods used in pre-
dicting SOM were based on traditional machine learning algorithms such as support
vector regression and random forest, as well as local methods such as LWR. However, the
modeling capability of these models for regional spectral libraries still requires improve-
ment [23,57,58]. The use of deep learning algorithms has shown great potential, and thus
needs to be further investigated.
A provincial-scale spectral library was utilized in this study, and the results show
that both the LSTM (R2 p = 0.83, RMSEp = 3.42 g/kg) and LSTM–CNN (R2 p = 0.96,
RMSEp = 1.66 g/kg) algorithms obtained higher accuracy when predicting SOM content
than the LWR model (R2 p = 0.82, RMSEp = 3.79 g/kg). Upon sorting the importance of
bands, it was discovered that the response bands for soil organic matter are primarily
Remote Sens. 2024, 16, 1256 11 of 13

located at the 800–900 nm, 1800–1900 nm, and 2200–2400 nm bands. Compared with
traditional statistical models, LSTM–CNN has stronger time-series modeling capabilities,
and can better capture correlations between SOM content and spectra, thus improving the
prediction accuracy.
By applying deep learning algorithms such as LSTM and LSTM–CNN on a provincial-
scale spectral library, it is possible to perform rapid, non-invasive, real-time monitoring
of soil properties within the study area. However, their performance on larger-scale SSLs
such as continental/global SSLs still needs to be further investigated. This study holds
significance in refining the method for SOM prediction, while also presenting a novel
approach to the use of deep learning in soil science. This advances the precision of basic
soil property prediction, and facilitates efficient agricultural land utilization.

Author Contributions: Conceptualization, W.J.; methodology, T.M. and W.J.; software, T.M., J.Y.
(Jianxin Yin) and Y.C.; validation, T.M. and J.Y. (Jianxin Yin); formal analysis, T.M. and W.J.; investiga-
tion, D.Y. and T.M.; resources, W.J.; data curation, T.M. and J.Y. (Jiajie Yang); writing—original draft
preparation, T.M.; writing—review and editing, W.J., B.L., X.Z., Y.H. and X.K.; visualization, T.M. and
J.Y. (Jianxin Yin); supervision, W.J. and B.L.; project administration, W.J.; funding acquisition, W.J.
and X.K. All authors have read and agreed to the published version of the manuscript.
Funding: This research is an output of Cropland Degradation Monitoring; Supported by Open Fund
of State Key Laboratory of Remote Sensing Science (Grant No. OFSLRSS202121); the National Natural
Science Foundation of China (42001048); Key Project of “Rejuvenating Mongolia with Science and
Technology” (NMKJXM202303); and the National Natural Science Foundation of China (42171289).
Data Availability Statement: The data presented in this study are available upon reasonable request
from the corresponding author.
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. Winfried, E.H.B. Functions of soil for society and the environment. Rev. Environ. Sci. Biotechnol. 2005, 4, 75–79.
2. Amundson, R.; Berhe, A.A.; Hopmans, J.W.; Olson, C.; Sztein, A.E.; Sparks, D.L. Soil and human security in the 21st century.
Science 2015, 348, 1261071. [CrossRef] [PubMed]
3. Jie, C.; Jing-Zhang, C.; Man-Zhi, T.; Zi-Tong, G. Soil degradation: A global problem endangering sustainable development. J.
Geogr. Sci. 2002, 12, 243–252. [CrossRef]
4. David, P. Soil Erosion: A Food and Environmental Threat. Environ. Dev. Sustain. 2006, 8, 119–137.
5. Evangelista, S.J.; Field, D.J.; McBratney, A.B.; Minasny, B.; Ng, W.; Padarian, J.; Dobarco, M.R.; Wadoux, A.M. A proposal for the
assessment of soil security: Soil functions, soil services and threats to soil. Soil Secur. 2023, 10, 100086. [CrossRef]
6. United Nations. The Sustainable Development Goals Report. 2022. Available online: https://unstats.un.org/sdgs/report/2022/
(accessed on 10 September 2022).
7. Sanchez, P.A.; Ahamed, S.; Carré, F.; Hartemink, A.E.; Hempel, J.; Huising, J.; Lagacherie, P.; McBratney, A.B.; McKenzie, N.G.;
Mendonça-Santos, M.D.; et al. Digital Soil Map of the World. Science 2009, 325, 680–681. [CrossRef]
8. Stenberg, B.; Rossel, R.A.; Mouazen, A.M.; Wetterlind, J. Visible and Near Infrared Spectroscopy in Soil Science. Adv. Agron. 2010,
107, 163–215.
9. Xu, Z.; Chen, S.; Lu, P.; Wang, Z.; Li, A.; Zeng, Q.; Chen, L. Optimizing a Standard Spectral Measurement Protocol to Enhance
the Quality of Soil Spectra: Exploration of Key Variables in Lab-Based VNIR-SWIR Spectral Measurement. Remote Sens. 2022,
14, 1558. [CrossRef]
10. Poppiel, R.R.; da Silveira Paiva, A.F.; Demattê, J.A.M. Bridging the gap between soil spectroscopy and traditional laboratory:
Insights for routine implementation. Geoderma 2022, 425, 116029. [CrossRef]
11. Shi, Z.; Ji, W.; Viscarra Rossel, R.A.; Chen, S.; Zhou, Y. Prediction of soil organic matter using a spatially constrained local partial
least squares regression and the Chinese vis–NIR spectral library. Eur. J. Soil Sci. 2015, 66, 679–687. [CrossRef]
12. Ji, W.; Viscarra Rossel, R.A.; Shi, Z. Improved estimates of organic carbon using proximally sensed vis-NIR spectra corrected by
piecewise direct standardization. Eur. J. Soil Sci. 2015, 66, 670–678. [CrossRef]
13. Nocita, M.; Stevens, A.; Noon, C.; Wesemael, B.V. Prediction of soil organic carbon for different levels of soil moisture using
Vis-NIR spectroscopy. Geoderma 2013, 199, 37–42. [CrossRef]
14. Stevens, A.; Nocita, M.; Tóth, G.; Montanarella, L.; van Wesemael, B. Prediction of Soil Organic Carbon at the European Scale by
Visible and Near InfraRed Reflectance Spectroscopy. PLoS ONE 2013, 8, e66409. [CrossRef] [PubMed]
15. Nocita, M.; Stevens, A.; Tóth, G.; Panagos, P.; Wesemael, B.V.; Montanarella, L. Prediction of soil organic carbon content by diffuse
reflectance spectroscopy using a local partial least square regression approach. Soil Biol. Biochem. 2014, 68, 337–347. [CrossRef]
Remote Sens. 2024, 16, 1256 12 of 13

16. Nawar, S.; Mouazen, A.M. On-line vis-NIR spectroscopy prediction of soil organic carbon using machine learning. Soil Tillage Res.
2019, 190, 120–127. [CrossRef]
17. Mendes, W.D.; Demattê, J.A.; Rosin, N.A.; Terra, F.D.; Poppiel, R.R.; Urbina-Salazar, D.; Boechat, C.L.; Silva, E.B.; Curi, N.;
Silva, S.H.; et al. The Brazilian soil Mid-infrared Spectral Library: The Power of the Fundamental Range. Geoderma 2022,
415, 115776. [CrossRef]
18. Moura-Bueno, J.M.; Dalmolin, R.S.; ten Caten, A.; Dotto, A.C.; Demattê, J.A. Stratification of a local VIS-NIR-SWIR spectral library
by homogeneity criteria yields more accurate soil organic carbon predictions. Geoderma 2019, 337, 565–581. [CrossRef]
19. Wijewardane, N.K.; Ge, Y.; Wills, S.; Libohova, Z. Predicting Physical and Chemical Properties of US Soils with a Mid-Infrared
Reflectance Spectral Library. Soil Sci. Soc. Am. J. 2018, 82, 722–731. [CrossRef]
20. Clairotte, M.; Grinand, C.; Kouakoua, E.; Thébault, A.; Saby, N.P.; Bernoux, M.; Barthès, B.G. National calibration of soil organic
carbon concentration using diffuse infrared reflectance spectroscopy. Geoderma 2016, 276, 41–52. [CrossRef]
21. Zhang, X.; Xue, J.; Xiao, Y.; Shi, Z.; Chen, S. Towards Optimal Variable Selection Methods for Soil Property Prediction Using a
Regional Soil Vis-NIR Spectral Library. Remote Sens. 2023, 15, 465. [CrossRef]
22. Araújo, S.R.; Wetterlind, J.; Demattê, J.A.; Stenberg, B. Improving the prediction performance of a large tropical vis-NIR
spectroscopic soil library from Brazil by clustering into smaller subsets or use of data mining calibration techniques. Eur. J. Soil
Sci. 2014, 65, 718–729. [CrossRef]
23. Shi, Z.; Wang, Q.; Peng, J.; Ji, W.; Liu, H.; Li, X.; Viscarra Rossel, R.A. Development of a national VNIR soil-spectral library for soil
classification and prediction of organic matter concentrations. Sci. China Earth Sci. 2014, 57, 1671–1680. [CrossRef]
24. Brodský, L.; Klement, A.; Penížek, V.; Kodešová, R.; Borůvka, L. Building soil spectral library of the Czech soils for quantitative
digital soil mapping. Soil Water Res. 2011, 6, 165–172. [CrossRef]
25. Brown, D.J.; Bricklemyer, R.S.; Miller, P.R. Validation requirements for diffuse reflectance soil characterization models with a case
study of VNIR soil C prediction in Montana. Geoderma 2005, 129, 251–267. [CrossRef]
26. Francos, N.; Chabrillat, S.; Tziolas, N.V.; Milewski, R.; Brell, M.; Samarinas, N.; Angelopoulou, T.; Tsakiridis, N.L.; Liakopoulos, V.;
Ruhtz, T.; et al. Estimation of water-infiltration rate in Mediterranean sandy soils using airborne hyperspectral sensors. Catena
2023, 233, 107476. [CrossRef]
27. Ramirez-Lopez, L.; Behrens, T.; Schmidt, K.; Rossel, R.A.; Demattê, J.A.; Scholten, T. Distance and similarity-search metrics for
use with soil vis—NIR spectra. Geoderma 2013, 199, 43–53. [CrossRef]
28. Rossel, R.A.; Behrens, T.; Ben-Dor, E.; Brown, D.J.; Demattê, J.A.; Shepherd, K.D.; Shi, Z.; Stenberg, B.; Stevens, A.; Adamchuk, V.I.;
et al. A global spectral library to characterize the world’s soil. Earth-Sci. Rev. 2016, 155, 198–230. [CrossRef]
29. Zhong, L.; Guo, X.; Xu, Z.; Ding, M. Soil properties: Their prediction and feature extraction from the LUCAS spectral library
using deep convolutional neural networks. Geoderma Int. J. Soil Sci. 2021, 402, 115366. [CrossRef]
30. Wang, S.; Guan, K.; Zhang, C.; Lee, D.; Margenot, A.J.; Ge, Y.; Peng, J.; Zhou, W.; Zhou, Q.; Huang, Y. Using soil library
hyperspectral reflectance and machine learning to predict soil organic carbon: Assessing potential of airborne and spaceborne
optical soil sensing. Remote Sens. Environ. 2022, 271, 112914. [CrossRef]
31. Ding, J.; Yang, A.; Wang, J.; Sagan, V.; Yu, D. Machine-learning-based quantitative estimation of soil organic carbon content by
VIS/NIR spectroscopy. PeerJ 2018, 6, e5714. [CrossRef]
32. Morellos, A.; Pantazi, X.E.; Moshou, D.; Alexandridis, T.K.; Whetton, R.L.; Tziotzios, G.; Wiebensohn, J.; Bill, R.; Mouazen, A.M.
Machine learning based prediction of soil total nitrogen, organic carbon and moisture content by using VIS-NIR spectroscopy.
Biosyst. Eng. 2016, 152, 104–116. [CrossRef]
33. Ji, W.; Li, S.; Chen, S.; Shi, Z.; Rossel, R.A.; Mouazen, A.M. Prediction of soil attributes using the Chinese soil spectral library and
standardized spectra recorded at field conditions. Soil Tillage Res. 2016, 155, 492–500. [CrossRef]
34. Ng, W.; Minasny, B.; Montazerolghaem, M.; Padarian, J.; Ferguson, R.; Bailey, S.; McBratney, A.B. Convolutional neural network
for simultaneous prediction of several soil properties using visible/near-infrared, mid-infrared, and their combined spectra.
Geoderma 2019, 352, 251–267. [CrossRef]
35. Ludwig, B.; Greenberg, I.; Vohland, M.; Michel, K. Optimised use of data fusion and memory-based learning with an Austrian
soil library for predictions with infrared data. Eur. J. Soil Sci. 2023, 74, e13394. [CrossRef]
36. Jia, X.; Xie, M.; Hu, B.; Zhou, Y.; Li, H.; Zhao, W.; Deng, W.; Shi, Z. Prediction of Soil Organic Carbon Contents in Tibet Using a
Visible Near-Infrared Spectral Library. Eurasian Soil Sci. 2023, 56, 727–737. [CrossRef]
37. Yang, M.; Chen, S.; Xu, D.; Hong, Y.; Li, S.; Peng, J.; Ji, W.; Guo, X.; Zhao, X.; Shi, Z. Strategies for predicting soil organic matter in
the field using the Chinese Vis-NIR soil spectral library. Geoderma 2023, 433, 116461. [CrossRef]
38. Rattan, L. Encyclopedia of Soil Science; CRC Press: Boca Raton, FL, USA, 2017.
39. Bao, S. Soil Agrochemical Analysis; China Agriculture Press: Beijing, China, 2000.
40. Haaland, D.M.; Thomas, E.V. Partial least-squares methods for spectral analyses. 1. Relation to other quantitative calibration
methods and the extraction of qualitative information. Anal. Chem. 2002, 60, 1193–1202. [CrossRef]
41. Steinier, J.; Termonia, Y.; Deltour, J. Smoothing and differentiation of data by simplified least square procedure. Anal. Chem. 1972,
44, 1906–1909. [CrossRef] [PubMed]
42. Krishnan, P.; Alexander, J.D.; Butler, B.J.; Hummel, J.W. Reflectance Technique for Predicting Soil Organic Matter 1. Soil Sci. Soc.
Am. J. 1980, 44, 1282–1285. [CrossRef]
Remote Sens. 2024, 16, 1256 13 of 13

43. Cleveland, W.S.; Devlin, S.J. Locally Weighted Regression: An Approach to Regression Analysis by Local Fitting. J. Am. Stat.
Assoc. 1988, 83, 596–610. [CrossRef]
44. Næs, T.; Isaksson, T. Locally Weighted Regression in Diffuse Near-Infrared Transmittance Spectroscopy. Appl. Spectrosc. 1992,
46, 34–43. [CrossRef]
45. Hochreiter, S.; Schmidhuber, J. Long short-term memory. Neural. Comput. 1997, 9, 1735–1780. [CrossRef]
46. Van Houdt, G.; Mosquera, C.; Nápoles, G. A review on the long short-term memory model. Artif. Intell. Rev. 2020, 53, 5929–5955.
[CrossRef]
47. Diederik, P.K.; Jimmy, B. Adam: A method for stochastic optimization. In Proceedings of the ICLR 2015, San Diego, CA, USA,
7–9 May 2015; CoRR, abs/1412.6980.
48. Agarap, A.F. Deep Learning using Rectified Linear Units (ReLU). arXiv 2018, arXiv:1803.08375.
49. Yang, L.; Cai, Y.; Zhang, L.; Guo, M.; Li, A.; Zhou, C. A deep learning method to predict soil organic carbon content at a regional
scale using satellite-based phenology variables. Int. J. Appl. Earth Obs. Geoinf. 2021, 102, 102428. [CrossRef]
50. Masters, D.; Luschi, C. Revisiting Small Batch Training for Deep Neural Networks. arXiv 2018, arXiv:1804.07612.
51. Sun, W.; Liu, S.; Zhang, X.; Li, Y. Estimation of soil organic matter content using selected spectral subset of hyperspectral data.
Geoderma 2022, 409, 13. [CrossRef]
52. Xu, B.B.; Li, D.C.; Shi, X.R. A Preliminary Study on Identification of Clay Minerals in Soils with Reference to Reflectance Spectra.
Pedosphere 1995, 5, 135–142.
53. Xu, Z.; Zhao, X.; Guo, X.; Guo, J. Deep Learning Application for Predicting Soil Organic Matter Content by VIS-NIR Spectroscopy.
Comput. Intell. Neurosci. 2019, 2019, 3563761. [CrossRef]
54. Yang, M.; Xu, D.; Chen, S.; Li, H.; Shi, Z. Evaluation of Machine Learning Approaches to Predict Soil Organic Matter and pH
Using vis-NIR Spectra. Sensors 2019, 19, 263. [CrossRef]
55. Zhang, L.; Cai, Y.; Huang, H.; Li, A.; Yang, L.; Zhou, C. A CNN-LSTM Model for Soil Organic Carbon Content Prediction with
Long Time Series of MODIS-Based Phenological Variables. Remote Sens. 2022, 14, 4441. [CrossRef]
56. Zhang, F.; Wu, S.; Liu, J.; Wang, C.; Guo, Z.; Xu, A.; Pan, K.; Pan, X. Predicting soil moisture content over partially vegetation
covered surfaces from hyperspectral data with deep learning. Soil Sci. Soc. Am. J. 2021, 85, 989–1001. [CrossRef]
57. Ramirez-Lopez, L.; Behrens, T.; Schmidt, K.; Stevens, A.; Demattê, J.A.; Scholten, T. The spectrum-based learner: A new local
approach for modeling soil vis–NIR spectra of complex datasets. Geoderma 2013, 195–196, 268–279. [CrossRef]
58. Shepherd, K.D.; Walsh, M.G. Development of Reflectance Spectral Libraries for Characterization of Soil Properties. Soil Sci. Soc.
Am. J. 2002, 66, 988–998. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.