Table of Content

20 May 2024, Volume 60 Issue 3

Previous Issue   

Multimodal Emotion Recognition Based on Hierarchical Fusion Strategy and Contextual Information Embedding

SUN Minglong, OUYANG Chunping, LIU Yongbin, REN Lin

2024, 60(3):  393-402.  DOI: 10.13209/j.0479-8023.2024.034

Asbtract ( )   HTML   PDF (5459KB) ( )  

Related Articles | Metrics

Existing fusion strategies often involve simple concatenation of modal features, disregarding personalized fusion requirements based on the characteristics of each modality. Additionally, solely considering the emotions of individual utterances in isolation, without accounting for their emotional states within the context, can lead to errors in emotion recognition. To address the aforementioned issues, this paper proposes a multimodal emotion recognition method based on a layered fusion strategy and the incorporation of contextual information. The method employs a layered fusion strategy, progressively integrating different modal features in a hierarchical manner to reduce noise interference from individual modalities and address inconsistencies in expression across different modalities. It leverages the contextual information to comprehensively analyze the emotional representation of each utterance within the context, enhancing overall emotion recognition performance. In binary emotion classification tasks, the proposed method achieves a 1.54% improvement in accuracy compared with the state-of-the-art (SOTA) model. In multi-class emotion recognition tasks, the F1 score is improved by 2.79% compared to SOTA model.

MFA-SGWNN: Botnet Detection Based on Multi-Feature Aggregation Spectral Graph Wavelet Neural Network

WU Hui, CHEN Xu, JING Yongjun, WANG Shuyang

2024, 60(3):  403-412.  DOI: 10.13209/j.0479-8023.2024.035

Asbtract ( )   HTML   PDF (995KB) ( )  

Related Articles | Metrics

In botnet attacks, because the characteristics of disguised botnet traffic data are too similar to normal traffic data, it is difficult to distinguish them accurately by traditional detection methods. In order to solve this problem, this paper proposes a Multi-feature Aggregation Spectral Graph Wavelet Neural Network (MFA-SGWNN). This method combines the attribute and spatial features of traffic, which can effectively capture the hidden characteristics of infected host traffic, enhance the feature representation of botnet nodes, and avoid the influence of unbalanced data samples and malicious encrypted traffic on detection. Experimental results on the ISCX2014 botnet and CIC-IDS 2017 (botnet) datasets show that MFA-SGWNN outperforms existing methods and has stronger robustness and generalization ability.

Few-shot Named Entity Recognition Method Based on Semantic Information Awareness of Labels

ZHANG Yue, WANG Changzheng, SU Xuefeng, YAN Zhichao, ZHANG Guangjun, SHAO Wenyuan, LI Ru

2024, 60(3):  413-421.  DOI: 10.13209/j.0479-8023.2024.036

Asbtract ( )   HTML   PDF (1177KB) ( )  

Related Articles | Metrics

Among various approaches of few-shot named entity recognition (NER), two-stage models based on prototype networks are widely used. However, these methods can not fully utilize the semantic information in entity labels and overly relies on entity type prototype vectors in distance calculation, resulting in poor generalization ability of the model. To address these issues, this paper proposes a few-shot named entity recognition method based on label semantic information awareness. This method consists of a two-stage process: entity span detection and entity type classification. When constructing entity type prototype vectors, the semantic information associated with the corresponding entity types is considered and fused with the prototype vectors through a dimension transformation layer. During the entity recognition of new samples, entity type positive and negative samples are combined with entity type prototype vectors to form entity type triplets, and the samples are classified based on the distance to the triplets. Experimental results on multiple datasets demonstrate that the proposed model significantly outperforms previous models.

Kinematics Analysis of Human Body Wearing a Hip-Joint-Assisted Exoskeleton

LIU Yu, HUANG Yan, ZHOU Zhihao

2024, 60(3):  422-430.  DOI: 10.13209/j.0479-8023.2024.032

Asbtract ( )   HTML   PDF (7569KB) ( )  

Related Articles | Metrics

Faced with the problem that existing studies on lower-limb powered exoskeleton lack joint-level kinematics analysis and assistance principle of hip exoskeleton, motion of human body wearing a hip assisted exoskeleton with multiple assist modes was recorded and analyzed. Motion data of human walking without exoskeleton, with transparent, low assistance mode, middle assistance mode, high assistance mode and resistance mode were obtained. Based on inverse kinematics and data processing, joint angles and gait characteristics were obtain. These results are helpful in clarifying the assistance principle of the hip exoskeleton, and may benefit design and motion control of assisted exoskeletons.

Changes in Wind Speed in China from 1981 to 2020 and Its Relationship with Atmospheric Circulation

ZHU Guoying, ZHAO Xinyi

2024, 60(3):  431-441.  DOI: 10.13209/j.0479-8023.2024.029

Asbtract ( )   HTML   PDF (1206KB) ( )  

Related Articles | Metrics

Based on the methods of piecewise regression model, Theil-Sen slope estimator, Mann-Kendall trend test, and stepwise regression model, this study explores the changes in near surface wind speed in China from 1981 to 2020 and its relationship with circulation patterns. The results indicate that overall, the wind speeds from 1981 to 2020 showed decreasing trends, with the decreasing trends of westerly, easterly, southerly, and northerly winds being −0.08, −0.06, −0.14 and −0.14 (m·s⁻¹)/10a, respectively. Using a segmented linear regression model to explore the inflection point of wind speed, the wind speed showed downwards trend before the inflection point and upward trends after the inflection point, and the average increase amplitude of southerly and northerly winds is higher than that of westerly and easterly winds. There are differences in wind speed changes in different regions and directions. The average south and north winds in northern China (Northwest, North, and Northeast China) changed from decreasing to increasing near 2012, while the average wind speeds in the four directions in southern China (South China) changed from decreasing to increasing near 1990. There is a strong correlation between the atmospheric circulation indices and the near surface wind speed in China. Among the inter-annual indices, the Asian Polar Vortex Intensity Index retains the highest frequency in the stepwise regression model, which is positively correlated with wind speed and has a higher impact on the northern sites of China than the central and southern sites. Among the interdecadal indices, the Pacific Decadal Oscillation retains the highest frequency in the stepwise regression model and is negatively correlated with wind speed. Its impact on wind speed at stations in southern China is significantly lower than that in the north and central regions

3-D S-wave Lithosphere Structure of South China Block Based on Ambient Noise Tomography

LÜ Jie, CHEN Yongshun, GUO Zhen

2024, 60(3):  442-452.  DOI: 10.13209/j.0479-8023.2024.018

Asbtract ( )   HTML   PDF (29507KB) ( )  

Related Articles | Metrics

Using continuous records of 704 stations from several mobile arrays, high-resolution S wave velocity structure of South China Block (SCB) was obtained by ambient noise tomography. In 3-D S wave velocity model, the velocities in shallow crust are well consistent with the surface geology and low velocity anomalies concentrate in Sichuan Basin and Jianghan Basin. The boundary between Yangtze Block and Cathaysian Block can be identified by a high-velocity belt in crust, which is Jiangnan Orogeny generated in collision. The low velocity anomaly of Cathaysian Block in middle crust is related to the widespread Mesozoic granite. In upper mantle, there are significant differences between the east and west South China Block: the west South China Block presents a significant high velocity anomaly which reflects a stable craton in Sichuan Basin; the low velocity anomaly in east South China Block reveals that the lithosphere has been thinned (<70 km) by the subduction of paleo-Pacific Plate. Besides, it is speculated that the low velocity anomaly of Qiongzhou Strait in upper mantle may correspond to the Hainan plume.

Prestack Surface Multiple Suppression Method Based on Matching Algorithm with Unsupervised Neural Network

LIU Lichao, HU Tianyue, LI Xixi, LIU Yimou, LIANG Shanglin, HUANG Jiandong

2024, 60(3):  453-463.  DOI: 10.13209/j.0479-8023.2024.031

Asbtract ( )   HTML   PDF (5488KB) ( )  

Related Articles | Metrics

To effectively suppress surface multiples in marine seismic data and then correctly image the exploration target, a prestack surface multiple attenuation algorithm based on unsupervised neural networks is proposed, which combines neural network methods with surface-related multiple elimination (SRME) methods. By continuously decreasing the learning rate, the unsupervised neural network replaces the matching filter operator for suppressing surface multiples in the prestack seismic data. This method requires neither traditional matching algorithms nor training on labeled datasets. The application of simple synthetic data, Sigsbee model data and field data verifies the effectiveness of the proposed method for surface multiple wave suppression.

Fluvial Geomorphic Characteristics of Taiwan Island, China and Its Tectonic Implications

GUO Xiaofei, LI Jianghai

2024, 60(3):  464-474.  DOI: 10.13209/j.0479-8023.2024.004

Asbtract ( )   HTML   PDF (20715KB) ( )  

Related Articles | Metrics

Based on SRTM DEM 30 m data and ArcGIS, the river system and geomorphological parameters (slope, topographic relief, and hypso-metric integral (HI)) were extracted to analyze the response of the river system for tectonic activities in Taiwan Island drainage basin, and the mainly conclusions were obtained as follows. 1) The rivers of Taiwan Island are controlled by tectonic activities, with the Central Range as the watershed. The rivers on the western side of Central Range flow vertically in the southern part of the system, and flow horizontally in the northern part. The rivers on the east of Central Range all flow in an east-west direction. 2) The HI values on Taiwan Island show an overall trend of high in the middle and low on the both sides, with a roughly positive correlation with elevation, and the HI values differ significantly along the east and west sides of the reverse fault boundary. 3) The HI values can reflect the generation time sequence of the orogenic belt of Taiwan Island. The longer the drainage basin has been affected by the tectonic uplift movement, the larger the HI value.

Fossil Record of Oscillatoriaceae from Cambrian Series 2 Xiaoerblaq Formation in Northwestern Tarim Basin, China

LI Jiayi, DONG Lin, SHI Kaibo

2024, 60(3):  475-488.  DOI: 10.13209/j.0479-8023.2024.016

Asbtract ( )   HTML   PDF (93409KB) ( )  

Related Articles | Metrics

Thin section observations have revealed a large amount of filamentous cyanobacteria from the dolomite of Cambrian Series 2 Xiaoerblaq Formation in the northwestern Tarim Basin. Specifically, four species belonging to three genera of Oscillatoriaceae fossils have been identified, including Obruchevella meishucunensis, Oscillatoriopsis longa, Oscillatoriopsis sp., and Siphonophycus typicum. These findings represent the first report of such filamentous cyanobacteria in the Xiaoerblaq Formation, and provide an exquisite preservation of their clear threelayer structure, that is, calcified mucilaginous sheath, organic tube wall, and internal black filament, augmenting the knowledge of their internal structure and identification characteristics. The discovery of Oscillatoriaceae fossils greatly increases the diversity of the Xiaoerblaq Biota and provides new materials for further study of fossils and the evolution of life in the early Cambrian.

Geometric and Kinematic Analysis of the Jiandingshan-Heiliangzi Anticlines, NW Qaidam Basin

PENG Luying, YANG Yizhou, ZHANG Junyong, WU Lei

2024, 60(3):  489-502.  DOI: 10.13209/j.0479-8023.2024.019

Asbtract ( )   HTML   PDF (1299KB) ( )  

Related Articles | Metrics

Jiandingshan and Heiliangzi anticlines in the Qaidam Basin, northern Tibetan Plateau are investigated to depict the intracontinental deformation pattern. Based on remote sensing images and seismic profiles, the structural patterns and kinematic characteristics of the surface faults and underground structures of the Jiandingshan-Heiliangzi anticlines are interpreted and analyzed. Remote sensing images show that most surface faults of the Jiandingshan anticline change from NW-SE striking transpressional to N-S striking normal faults to the southeast, while most surface faults of the Heiliangzi anticline are NNE striking sinistral transtensional faults. Seismic profiles show that the Jiandingshan-Heiliangzi anticlines are composed of upper thrust detachment fault systems and lower basementinvolved positive flower structures. In addition, it is inferred that the decoupling between shallow and deep structures of the Jiandingshan and Heiliangzi anticlines is influenced by the gypsum-salt layers in Lulehe Formation. Considering the fault map, kinematic features, and along-strike variations of the surface faults and basement-involved faults, the Jiandingshan-Heiliangzi anticlines are treated as dextral transpressional structures. It is inferred that the interior of the Qaidam Basin has experienced differential eastward extrusion during its continuous shortening induced by the convergence of the Indian and Eurasian continents, leading to a series of transpressional anticlines in the basin.

Geomagnetic Field Paleointensity Study on Archaeological Pottery from Xinjiang Uygur Autonomous Region, China

LI Nengtao, ZHOU Tinghong, HUANG Baochun

2024, 60(3):  503-512.  DOI: 10.13209/j.0479-8023.2024.021

Asbtract ( )   HTML   PDF (14565KB) ( )  

Related Articles | Metrics

A geomagnetic field paleointensity study using an improved double heating Thellier-Thellier type method (IZZI method) and a rock magnetic study were conducted on archaeological pottery samples from Changji, Kashgar, Tashkurgan, and other areas in Xinjiang Uyger Autonomous Region (China). The results show that some samples carried stable primary remanent magnetization, from which geomagnetic field paleointensity data between ≈1000BC and ≈1200AD were acquired. Series criteria were applied to select reliable result and corrections were performed to the paleointensity data. Virtual axial dipole moments (VADM) of ≈5×10²²–11×10²² Am² were obtained. The paleointensities obtained are roughly in the same range as the data published in other regions of China in the same period, but the amplitude of variation is significant, which is inconsistent with the global geomagnetic field model SHAWQ2k. The new data were compared with the geomagnetic field intensity peak in the Middle East region at ≈800AD and the geomagnetic field intensity peak in East Asia region at ≈200AD. The fitting results indicate a geomagnetic field intensity peak might exist in Xinjiang region at ≈540AD, with the time of reaching the peak between East Asia and the Middle East region. This result is consistent with the westward drift characteristics of non-dipole fields, further indicating that this geomagnetic field paleointensity peak may be related to the dynamics of the Earth’s outer core.

Supplemental Evaluation of Landslide Displacement Prediction Models

TIAN Yuan, CHANG Xiaoyin, ZHAO Wenyi, CHENG Chuyun, Bronte Scheuer, DENG Yanglanduo, MA Ruiping, ZHANG Jianxue

2024, 60(3):  513-520.  DOI: 10.13209/j.0479-8023.2024.022

Asbtract ( )   HTML   PDF (834KB) ( )  

Related Articles | Metrics

The distinguishing ability of typical overall indicators is becoming increasingly insufficient in the evaluation tasks of landslide displacement prediction models while the predicting ability of the models in the highrisk deformation periods has unexpectedly not been paid enough attention. Therefore, two supplemental evaluation indicators for landslide displacement prediction models, combined error of rapid displacement period (CERDP) and peak prediction error (PPE), along with their application process, are accordingly designed. To verify the feasibility and performance of the proposed supplemental evaluation indicators, a case study based on actual landslide monitoring data has been carried out. It can be concluded that the proposed supplemental evaluation indicators and application process are reasonable, feasible, and practical for effectively addressing the evaluation issues caused by insignificant differences in indicators, such as root mean squared error (RMSE), and are able to finally select an overall reliable model which performs best during rapid displacement periods and around peak deformation times, thus providing effective technical support for landslide risk management tasks.

High Precision and Dense Observation of High-Speed Rail Seismic Wavefield and Its Application Prospects: Observation and Processing Results Based on Communication Optical Cable

HE Xiangge, SHI Yongxiang, SU Qingqing, ZHANG Min, LU Hailong

2024, 60(3):  521-527.  DOI: 10.13209/j.0479-8023.2024.033

Asbtract ( )   HTML   PDF (13675KB) ( )  

Related Articles | Metrics

The high-precision distributed acquisition of high-speed rail seismic waves were achieved using the High-Definition Distributed Acoustic Sensing (HD-DAS) system. The sensor unit used simple communication optical cable to achieve good monitoring effects at low cost, and could be easily deployed according to actual environmental conditions, greatly improving the efficiency of field observation work. By using multiple parallel optical cables, noise could be effectively suppressed and the signal-to-noise ratio of high-speed rail seismic signals can be improved. Data analysis and processing were performed on seismic signals collected by HD-DAS from high-speed rail. The analysis revealed that the signals display discrete spectral characteristics. Further analysis was conducted on the attenuation law of seismic wave signals, and the transmission speed of seismic waves near the surface was obtained through the frequency-wavenumber spectrum. This indicated that the high-speed rail seismic wave signals acquired through optical fiber could be used to detect underground structures and monitor changes in the medium’s state, thereby enhancing the safety of high-speed rail operations.

Impacts of Production-Living-Ecological Space Transformation in the Heihe River Basin on Ecological Environment Quality in the Context of Climate Change

ZOU Yi, ZHAO Yiyang, MENG Jijun

2024, 60(3):  528-538.  DOI: 10.13209/j.0479-8023.2024.009

Asbtract ( )   HTML   PDF (3089KB) ( )  

Related Articles | Metrics

Choosing the Heihe River Basin in the arid northwest of China as the study area, using ArcGIS 10.6 and Mathematica 8.0 as research platforms, the paper analyzes the spatio-temporal pattern of the transformation of the production-living-ecological spaces from 1990 to 2020. Combining the vegetation index (NDVI) and salinity index (SI), we investigate the spatiotemporal changes of eco-environmental quality and separate the contributions of explicit and implicit transformation of the production-living-ecological spaces to changes in eco-environmental quality by establishing a mixed effects model and attribution analysis model under the background of climate change. The results show that 1) the strong and weak ecological spaces in the Heihe River Basin account for 28.12% and 65% respectively, while the living and production space accounts for only 6.88%. The living and production spaces have each increased by 0.32% and 0.96% over the past 30 years, and frequent transformations occur between strong and weak ecological spaces (72.8%). The change in eco-environmental quality has transitioned from a decline to a sustained improvement trend since 2000. 2) The contributions of explicit transformation of the production-livingecological spaces and climate change to eco-environmental quality are 39.8% and 36.6%. Precipitation has the main contribution to the ecological environment of production and living spaces, while temperature has a nonlinear impact on the eco-environmental quality of ecological spaces. The contribution rate of implicit transformation to ecological environment changes is 23.6%. 3) The transformation of the three spatial sectors has a stronger spatial heterogeneity effect on ecological environment, while the temporal heterogeneity effect of climate change is stronger. The eco-environmental quality in the middle reaches of the river basin is simultaneously affected by explicit and implicit transformation, while the downstream is mainly affected by implicit transformation. The results provide a reliable theoretical basis for optimizing the land spatial development and ecosystem management in the Heihe River Basin.

Pathways and Mechanisms of Nitrogen Removal by Hydrogen-Oxidizing Bacteria under Stress of Quinolone Antibiotics

WANG Xisong, SHI Gang, DING Lingyun, ZHANG Yaoyu, ZHANG Shanfa, Maurycy Daroch, TAO Huchun, ZHANG Lijuan

2024, 60(3):  539-547.  DOI: 10.13209/j.0479-8023.2024.010

Asbtract ( )   HTML   PDF (2342KB) ( )  

Related Articles | Metrics

A hydrogen-oxidizing bacterium Ideonella sp. TH17 was investigated for ammonium removal from wastewater. Possible pathways and mechanisms of nitrogen assimilation under the stress of two typical quinolone antibiotics and their mixture were studied by molecular biological approaches. The results showed that the capability of ammonium assimilation by TH17 was stimulated at lower concentration (5 μg/L), but inhibited at elevated concentration of quinolone antibiotics. The inhibitory effect of mixed antibiotics was weaker than that of individual antibiotics at higher concentration. Under stress of 5 μg/L quinolone antibiotics, the expression of ammonium assimilation genes was up-regulated, and activities of functional and antioxidant enzymes were enhanced. Meanwhile, the expression of antibiotic resistance gene was up-regulated, which increased the tolerance of TH17 to quinolone antibiotics. Further, amino acid, purine and biotin metabolisms were up-regulated significantly to provide energy, substrates, and coenzymes for enhanced nitrogen metabolism of TH17. Therefore, the influencing pathways and mechanisms of low-concentration quinolone antibiotics on nitrogen removal by TH17 were revealed by gene, enzyme and metabolite molecules, which could guide the development and application of novel processes for biological nitrogen removal from wastewater.

Variation of Concentration, Form and Flux of Phosphorus in the Three Gorges Reservoir from 1998 to 2019

LOU Baofeng, YU Mingxing, HUANG Bo, YANG Yan, LIN Jingjing

2024, 60(3):  548-562.  DOI: 10.13209/j.0479-8023.2024.027

Asbtract ( )   HTML   PDF (1356KB) ( )  

Related Articles | Metrics

The changes in total phosphorus (TP) concentration and form in the Three Gorges Reservoir (TGR) from 1998 to 2019 were studied, and P fluxes were calculated. The entire research period was divided into three stages: 1998–2002 (stage I, before the operation of the TGR), 2003–2012 (stage II, as a transitional period), and 2013–2019 (stage III, after the TGR achieved the normal water level of 175 m and Xiangjiaba Reservoir began to operate). Comparative analysis was mainly on Stages I and III. The results show that the concentrations of suspended sediment (SS) in the TGR decreased to a very low level (average 42 mg/L during 2013–2019 period and 96 mg/L in wet season). From 1998 to 2019, the inflow and outflow of sediment amount showed a downward trend. A huge amount of sediment was retained at the bottom of the reservoir during 2003–2012 (average 154.6 million t/a), while in 2013–2019, the amount of sediment deposition significantly decreased and the sedimentation amount was only 37% of that in 2003–2012. Year 2013 marked a turning point in the trend of phosphorus pollution in TGR from increasing to decreasing. Afterwards, concentrations of TP and DP (dissolved phosphorus) showed a significant downward trend. The form of phosphorus in TGR changed from mainly granular to dissolved, and the percentage of DP in TP in the perennial backwater area increased from 39% before TGR formation to 69% during 2017–2019. 2004–2012 was a period with high deposition for TP when the amount of TP entering the reservoir was large and fluctuated vastly between years, and approximately 60% of phosphorus was trapped (82220 t/a). From 2013 to 2019, there occurred significantly decrease in the TP amount entering the reservoir in accompany with diminishing interannual fluctuation, tending to a balance between input and output of P. Fluxes of TP, DP, and PP (particulate phosphorus) all show a pattern of wet season>even season>dry season. The main driving forces for changes in TP concentration, flux, and form are the reduction in SS concentration and phosphorus pollution level.

Characterization and Application of an Online Measurement System for NO₃ and N₂O₅ Based on Cavity Ring-Down Spectroscopy

ZHANG Tiantian, ZUO Peng, MA Jingyu, YE Chunxiang, LIN Weili, ZHU Tong

2024, 60(3):  563-574.  DOI: 10.13209/j.0479-8023.2024.030

Asbtract ( )   HTML   PDF (4284KB) ( )  

Related Articles | Metrics

This study presents a home-made online measurement system for atmospheric NO₃/N₂O₅ based on cavity ring-down spectroscopy and its characterization results. The system achieves a detection limit of 1.7×10⁻¹² for NO₃ and 2.9×10⁻¹² for N₂O₅ at a time resolution of 1 minute. Under sampling conditions of 6 L/min flow rate and a filter membrane used for 1 day, the sampling loss for NO₃ and N₂O₅ is determined to be 17.0% and 18.0%, respectively. Allan variance analysis indicates an optimal system stability at integration times between 60 and 100 seconds. Field measurements of NO₃ and N₂O₅ were conducted in the urban area of Beijing, and the results demonstrate that the measurement system can achieve high-sensitivity measurements of NO₃ and N₂O₅ under atmospheric environment.

Analysis of Ecosystem Carbon Footprint for Three Staple Crop Farmlands in China

LI Yaning, WU Xiuqin

2024, 60(3):  575-584.  DOI: 10.13209/j.0479-8023.2024.026

Asbtract ( )   HTML   PDF (1849KB) ( )  

Related Articles | Metrics

Based on the China net primary productivity data of vegetation from 2006 to 2019, the distribution data of three staple crops, and agricultural production inputs, this article estimates the carbon sink, agricultural carbon emissions, and carbon footprint of China’s three staple crops, and analyze the spatial and temporal patterns of carbon sinks in the three staple crops. The results show that 1) from 2006 to 2019, the total carbon sequestration of the three staple crops in China continues to increase, with rice, maize, and wheat accounting for 41.92%, 38.62%, and 19.46%, respectively. The carbon sequestration capacity of three staple crops far exceeds the carbon emissions caused by agricultural production factors, with a ratio of 5.37:1, indicating that the China agricultural ecosystem has a strong carbon sequestration capacity. 2) The total agricultural carbon emissions in China have continued to increase, but the growth rate has decreased. Fertilizer and diesel are the main carbon sources. 3) The ecosystem of three staple crop farmlands shows a significant carbon ecological surplus. The carbon footprint in 2019 is 17.55×10⁶ hm², accounting for 18.62%, which means that the CO₂ emissions from the production of the three staple crops farmland system are reduced to less than 1/5 of the China three staple farmland to absorb. The results can help clarify their position in the carbon cycle and provide important scientific basis for promoting agricultural carbon emission reduction, achieving “carbon peak” in 2030 and “carbon neutral” in 2060.