To address insufficient feature extraction due to individual deep learning models’ limitations in capturing local and global features, and the tendency to underestimate high yields and overestimate low yields, a new deep learning model called convolutional neural network (CNN)-transformer with serial connection (CNN-transformers) was introduced to estimate winter wheat yield by combining the local feature extraction strengths of CNNs with the global information extraction abilities of transformer networks utilizing self-attention mechanisms. The remote sensing technology included temperature and spectral response indicators; the vegetation temperature condition index (VTCI), leaf area index (LAI), and fraction of photosynthetically active radiation (FPAR) aggregated over ten-day periods. Compared to the CNN model, the transformer model, the CNN-transformer model with parallel connection (CNN-transformerp), and the transformer-CNN model with serial connection (transformer-CNNs), the CNN-transformers achieved higher accuracy in estimating winter wheat yield (${R} ^{2} =0.70$, RMSE =420.39 kg/ha, MAPE =7.65%), which was capable of extracting more information related to yield from various remotely sensed parameters and addressing the problems of high yield underestimation and low yield overestimation observed. The robustness and generalization of the CNN-transformers were further assessed through the fivefold cross-validation and the leave-one-year-out methods. In addition, utilizing the CNN-transformers, the study uncovered the cumulative impact of the winter wheat growth period, examined how incrementally adding data at ten-day intervals affects yield estimation, and assessed the model’s proficiency in depicting growth accumulation during the growth process. The findings indicated that the model well identified the crucial growth phase of winter wheat between late March and early May.

Accurate crop yield estimation in advance is of paramount importance for making appropriate decisions in adjusting food price and formulating food policy. Most traditional statistical yield estimation models directly capture nonlinear relationship between the original time series remotely sensed variables and yields, ignoring the complexity and non-stationarity in the remotely sensed variables. This study integrated variational mode decomposition (VMD), correlation analysis and neural network architecture to construct a Decomposition-Screening-Reconstruction-Estimation paradigm for wheat yield estimation. Firstly, VMD optimized by whale optimization algorithm (WOA) was applied to decompose the time series remotely sensed variables into intrinsic mode function (IMF) components. Secondly, effective IMF components with strong correlation to the original time series remotely sensed variables were identified and reconstructed as the key time series features. Finally, the performances under the proposed paradigm of Transformer model and baseline models including support vector regression (SVR), random forest (RF) and long-short term memory (LSTM) were compared. The results showed that WOA-VMD-Transformer (WVTF) (R2= 0.66, RMSE = 459.09 kg/ha, MRE = 8.22 %) outperformed other models, and deep learning models (LSTM and Transformer) based on the new paradigm exhibited superior performance compared with traditional models (SVR and RF). In addition, WVTF (R2=0.54, RMSE=468.19 kg/ha, MRE=6.13 %) provided better performance at the sampling sites compared with previous proposed model (R2= 0.45, RMSE = 738.63 kg/ha, MRE = 8.21 %). The estimated wheat yields of 2019-2024 based on the optimal model were identical to the distribution of actual yields. In conclusion, the framework for yield estimation under novel paradigm extracts the original remotely sensed variables into key time series features, which provides an important reference for regional crop yield estimation and agricultural development.

Precise crop yield estimation is crucial for maintaining national food security. Crop growth models, statistical regression methods, and deep learning methods are among the numerous methods available for estimating crop yields. Deep learning, with its robust feature extraction capability and data-driven features, has propelled agriculture forward by leaps and bounds, and one of the applications oriented to national needs is the agricultural application of deep learning-based AI technology. However, as research and applications deepen and grow, the limitations of data-driven methodologies have emerged, such as difficulties in migrating reuse, reliance on samples, and poor interpretability. The organic integration of knowledge in various forms of expression and deep learning is the way to realize the dual drive of knowledge and data, and the integration of the two has not been well investigated. Therefore, we proposed a knowledge-guided deep learning (KGDL) framework that integrated knowledge with CNN and BiLSTM networks from both feature-level and network-level perspectives for winter wheat yield estimate in the Guanzhong Plain from 2007 to 2021. Our results showed that incorporating feature-level and network-level knowledge significantly enhanced wheat yield estimates, achieving R2= 0.73, RMSE = 447.12 kg/ha, and MAPE = 8.72%, reducing uncertainty by 109.88 kg/ha for the high-yield dataset, 81.73 kg/ha for the medium-yield dataset and 55.34 kg/ha for the low-yield dataset. Furthermore, the KGDL model, driven by both knowledge and data, resulted in a gradual increase in the coefficients between the output and the target yield on a layer-by-layer basis, with the average coefficients increasing from 0.35 to 0.58. Overall, the proposed framework shows promising application prospects for deep learning-based crop yield estimation, which can provide important technical support for promoting precision production in regional agriculture.

Land surface temperature (LST) is a critical physical parameter affecting energy and water exchange that has attracted much attention in various fields, such as environmental protection, agriculture, and climate change. Studies on spatially continuous and high-resolution LST retrieval methods, which can be efficiently acquired using thermal infrared (TIR) remote sensing technology, have been developed for many years, resulting in various LST remote sensing products. The typical mechanism thermal radiative transfer model is based on the assumption that the land surface is flat, with the TIR remote sensing image of the spatial resolution of the enhancement of the ability to observe the land surface of the three-dimensional geometric structure of the fine observation, due to the terrain caused by the topographic effect caused by the topography of the undulation becomes non-negligible, the assumption of flat surface may cause apparent errors. Some LST retrieval algorithms considering topographic effects have also been proposed recently. However, they are still inaccessible due to dependence on emissivity or atmospheric parameters, which limits the accuracy and timeliness of the retrieval algorithms. In addition, various machine learning algorithms for end-to-end LST retrieval have been proposed, which utilize their ability to handle complex nonlinear relationships to retrieve LST without external parameters. However such models currently do not fully consider the topographic effect due to a lack of account of the radiative transfer process in undulating terrain conditions. In this study, utilizing the ability of convolutional neural networks to extract spatial features from adjacent pixels, a radiative transfer model-driven convolutional neural network (CNN) model is proposed to realize the end-to-end retrieval of LST, considering the topographic effect. During training, a computational method based on ambient radiance scattered from the surrounding adjacent pixels in the improved radiative transfer model is used to obtain a local-scale simulation dataset covering different LSTs, emissivity, terrain undulations, and atmospheric conditions. The proposed CNN model is trained on this basis, and the theoretical accuracy is evaluated using the simulation dataset. The model has been applied to long-time-series Landsat-9 TIR remote sensing images. The accuracy is verified using terrain-corrected (TC) LST products. The results show that the new method proposed in this paper can effectively eliminate the topographic effect in TIR remote sensing observations and obtain accurate LST retrieval results, requiring only brightness temperature and digital surface model data.

Urban land surface temperature (ULST) is one of the core parameters in monitoring the urban thermal environment, which has received extensive attention in several study and application areas. Thermal infrared (TIR) remote sensing technology can efficiently observe large-scale land surface thermal radiance information and is a critical approach used to obtain ULST quickly. Traditional LST retrieval algorithms are conducted using the classical radiance transfer equation (RTE) based on the assumption that the land surface is flat, which may be challenging to hold for complex urban landscapes. Moreover, with the improvement of the spatial resolution of remote sensing images, the influence caused by the geometric structure will be more obvious. Various urban thermal radiance transfer models have been proposed and successfully applied to TIR remote sensing images with tens of meters spatial resolutions, such as Landsat, ECOSTRESS, and Gaofen-5. Current airborne TIR sensors can observe remote sensing images with ultra-high spatial resolution (sub-meter). In this paper, using the ensemble learning method based on the ultra-high spatial resolution urban thermal radiance transfer model (UHURT), a new retrieval algorithm is developed to estimate the LST directly from the observed brightness temperature. The proposed new algorithm applies to ultra-high spatial resolution remote sensing images. It has the end-to-end advantage of not relying on atmospheric parameters or land surface emissivity, known as in traditional algorithms, thus avoiding the limitations due to the lack of available input data. Validation results based on the simulation dataset showed that the proposed algorithm has higher theoretical accuracy than the traditional split-window algorithm. As the sky view factor (SVF) decreases, the accuracy advantage becomes more pronounced, growing from 0.149 K (SVF = 1.0) to 1.085 K (SVF = 0.25). The application results in the remote sensing image also indicated that the results of the proposed algorithm (RMSE = 2.093 K) are more accurate than those of the SW algorithm (RMSE = 2.490 K), and the correlation between the resultant error and building density is lower, which can accurately reduce the geometric effect to obtain the ULST better.

Thermal infrared (TIR) remote sensing is an important technological tool for observing large-scale land surface thermal radiance and can obtain the spatially continuous land surface temperature (LST), a critical land surface physical parameter of great interest in several fields. After decades of development, various LST retrieval algorithms have been proposed. However, current studies indicated that the commonly used retrieval algorithms show a decrease in the accuracy of the results under humid atmospheric conditions, and the theoretical analysis of the phenomenon needs to be developed. This study derives the LST error as a function of atmospheric parameters (transmittance, upward radiance, and downward radiance) directly based on the TIR radiative transfer equation. Compared with the TIR channel, the mid-infrared (MIR) channel has less water vapor absorption, is more insensitive to water vapor, and has a larger transmittance, which is expected to improve the accuracy of LST retrieval under humid atmospheric conditions. In this study, a typical simulation dataset under various atmospheric and land surface conditions is constructed based on the MIR channels of MODIS remote sensing data. Analysis of the retrieval results based on the simulation datasets shows that the MIR channels have more minor errors for the same level of atmospheric errors. With the growth of column water vapor (CWV), the error of the split-window (SW) algorithm constructed based on the TIR channel increases. In contrast, the accuracy of the algorithm developed by MIR channels is more stable, and the advantage of the accuracy in humid atmospheric conditions is more prominent. Two SW algorithms are applied to nighttime TIR and MIR remote sensing images observed by Aqua MODIS, and the validation results obtained based on SURFRAD ground sites also showed that the two MIR SW algorithms achieved the accuracy advantage of 0.425 K (SW1_TIR: 2.582 K / SW1_MIR: 2.157 K) and 0.525 K (SW2_TIR: 2.624 K / SW2_MIR: 2.099 K), which indicated that the MIR-SW algorithms can more accurately retrieve the LST under humid atmospheric conditions.

Due to climate change, the impact of drought on field crop production is extremely important. This study focuses on the vegetation temperature condition index (VTCI), an index-based drought monitoring index that can characterize drought conditions in near real time (at ten-day intervals), and explores the applicability of different spatial and temporal data fusion schemes to it. It also proposes a field-scale VTCI fusion framework based on the Sentinel-3 VTCI calculation and the land surface temperature (LST) downscaling. First, based on analyzing the computational characteristics of VTCI, multiyear VTCI based on Sentinel data sources was obtained, which further expands the diversity of data sources for VTCI. On this basis, a combination of qualitative and quantitative methods was used to compare the applicability of two schemes: Scheme 1, based on the “blend-then-index” (BI) strategy, which first fuses normalized difference vegetation index (NDVI) and LST, and then calculated the fused VTCIs; and Scheme 2, based on the “index-then-blend” (IB) strategy, which directly fuses the VTCIs based on the calculated VTCIs. It was found that all the fused VTCIs remained highly correlated with the ten-day cumulative precipitation. Compared with the fused VTCIs obtained by Scheme 2, the VTCIs obtained by Scheme 1 were able to display more spatial details. In addition, the VTCIs of Scheme 1 were more consistent with the Sentinel-3 VTCIs, and the accuracy of field yield estimation using the fused VTCIs was higher ($r$of 0.58 and root-mean-square error (RMSE) of 783.27 kg/ha).

The land surface temperature (LST) of urban is a key factor in the field of urban environmental monitoring, and thermal infrared (TIR) remote sensing is an efficient method to obtain it. An important assumption of the traditional thermal radiance transfer model is that the land surface is flat, which has now proven difficult to hold under the urban landscape. Most of the existing urban thermal radiance transfer models have been developed for remote sensing images with a spatial resolution of tens of meters. Currently, airborne TIR sensors have the observation capability to acquire remote sensing images with an ultra-high spatial resolution (1 cm to 1 m), and the model needs to be improved. This paper proposed a new ultra-high spatial resolution urban thermal radiance transfer model (UHURT) after analyzing the transfer processes within the urban canopy at ultra-high spatial resolution. Various radiance components, the emitted radiance, reflected atmospheric downward radiance, and adjacent radiance, were modeled separately. The results of the traditional model and the UHURT model were compared with the results of a ray-tracing computer simulation model, which showed that the new model successfully quantifies the multiple scattering and adjacent effects, and obtained images closer to the computer simulation images. Besides, the LST retrieval of the computer-simulated images was performed using the traditional model and the UHURT model, and the proposed model successfully reduced the errors of the retrieval results and weakened the spatial correlation between the residual distribution and the geometric characteristics of the urban landscape.

As a critical parameter of the land surface energy balance, land surface temperature (LST) has received extensive attention in various fields. Thermal infrared (TIR) remote sensing can efficiently obtain large-scale, long-time-series information on land surface thermal radiance. Multiple algorithms, including physics-based and deep learning algorithms, have been proposed for obtaining LST from the observations. In algorithm analysis, the theoretical performance is typically evaluated through simulated data representing atmospheric and land surface environmental conditions that vary globally before applying it to authentic TIR remote sensing images and validating the accuracy using ground-measured data. However, due to the complexity of the observation environment, errors in results obtained from validation using ground-measured data tend to be larger than theoretical errors obtained using simulated data, which limits the performance of the algorithm in practical applications. Obtaining globally covered, representative ground-measured data and synchronized observations of remote sensing images are costly, making it difficult to provide enough data to develop new algorithms. The transfer learning method can learn from the pre-trained deep learning model that applies to the source task, and fine-tuning using only a small number of samples from the target task can result in a well-performing model. This paper proposes a modified transfer-learning-based (TL) land surface temperature retrieval algorithm to pre-train a knowledge-driven deep neural network model using a simulation dataset and then use a small amount of ground-measured data for tuning to obtain the final LST retrieval model. The new algorithm was applied to the remote sensing data observed by the Landsat-8 Thermal Infrared Sensor (TIRS), and the validation results based on the ground-measured data and global Landsat-8 LST products showed that the RMSE of the fine-tuned model result was about 0.4 K lower than the Landsat-8 product, and about 0.3 K lower than the pre-trained model results, reaching 2.2 K. Moreover, the results are in good agreement with the Landsat LST product in multiple regions worldwide with different land cover types, which demonstrated the effectiveness and stability of the proposed TL algorithm.

This paper gives a brief description of an Airborne Multi-angle Power Line Inspection System (AMPLIS). AMPLIS is composed by 3 CCD cameras, a Position and Orientation System (POS), a stabilized platform, the data collection and control subsystem. It can be equipped on a helicopter and fly along the lines at a speed of about lOOkm/h at a relative height of 100 m over the power lines. AMPLIS is capable of detecting the distance between the power lines and the ground surface with an accuracy of less than 0.5 m. It can automatically find the dangerous objects beneath the lines which can greatly decrease the man power and cost in power line inspection. It has been successfully tested with good performance in Wuhan, China, 2005.

Thermal inert in method is one of the main approaches for drought monitoring by using remotely sensed data. The method is widely used in the China North Plain which is relatively flat. In our study, we tried to apply the approach to the agricultural areas of Shaanxi Province in the Northwest China with terrain and climate varies. The results showed that apparent thermal inertia can be used to monitor the drought occurrence. Considering the effects of terrains, vegetation coverage and soil types, the more homogeneous the land surface is, the better correlation between the apparent thermal inertia mid surface soil moisture is

A modified diurnal land surface temperature change model was developed for soil thermal inertia retrieval by using NOAA/AVHRR remotely sensed data. The modified thermal inertia model was used to estimate surface soil moisture contents in the Guanzhong Plain of Shaanxi Province in the Northwest China. The results showed that the retrieved values of soil thermal inertia converged when the Fourier series were set to 10 or greater than 10, and the values were in the range of ground measured values published in some related articles. For applications of the model, soil thermal inertia can be reversed by applying the second Fourier series approximation. Based on the significance and the range of the estimated surface soil moisture, we found the exponential model between soil thermal inertia and soil moisture had the best performance in estimating soil moisture contents

One of the problems in normalized difference vegetation index (NDVI) based agricultural applications is the time lag between precipitation and NDVI. In this study, the CERES-Wheat model under decision support system for agrotechnology transfer shell was used to simulate the time lag between precipitation and NDVI under minted conditions in the Guanzhong Plain, China. The results showed that there were about 14 to 37 days' time lags between precipitation and NDVI, and the time lags depended on the growing stages of winter wheat. The time lag was about 14 days at the crop's tasselling stage, while it was about 37 days for the crop's reviving stage. The results also indicated that there were year to year variations of the time lags and the time lag should be considered for NDVI based agricultural applications.

This paper mainly proposed using multi-temporal spectral data with given weight to estimate yield using contribution ratios of different stages to improve yield estimation, then use stepwise regression to build models for yield estimation. The contribution ratio is calculated by principal component analysis respectively. Result show that this methodology reflects the crop growth status, physiological characters at different stages, and improves the accuracy obviously.

One problem in quantitative remote sensing inversion is the correlations between variables. Path analysis is a statistical technique that differentiates between correlation and causation, features multiple linear regressions, and generates path coefficients. In this paper, path analysis was applied to study the correlations and causations of two cases in remote sensing reversion. One is the retrieval of land surface temperature, and another is to explain the results of land surface moisture estimation. We found that path analysis can be used to study the direct effect and the indirect effects of a variable in remote sensing inversion, and gave a better explanation of the results of multiple linear regression analysis.

At XiaoTangShan Precision Agriculture Experiment Base, suburban of Beijing city, we obtained the hyperspectral data of wheat canopy by field measurement and synchronal relative foliage water content (RFWC) at lab, according to the different growth stage of wheat, in the spring of 2002. In this paper, we extracted spectral feature parameters firstly. Then, the correlation analysis between the spectral features parameters and RFWC was made. two spectral feature parameters, which have good relativity to foliage water content, to make linear regression models between RFWC and spectral feature parameter according to the growth stages of wheat.

LAI (Leaf Area Index) is an important parameter for plant growth status detecting. The traditional method to measure leaves' area in field is time and labor consuming. In some remote sensing applications, LAI might be needed more intensively such as daily and weekly. Crop simulation models can be effective tools for simulating LAI, crop and soil water statuses, crop growth and development parameters at the daily step. Ground LAI measurements for winter wheat were carried out in several fields of Shunyi county, Beijing, China at several days' interval during the crop growth season. The CERES-wheat model is run under local soil, weather and management conditions of a field site to simulate daily LAI values. The simulated LAI corresponded comparatively well with the measured ones at the early stage. The results suggested that the technique might be promising for estimating LAI, this make it possible to develop an approach to assimilate LAI of winter wheat between ground measurements and CERES-wheat simulation.