Search

Fusing multi-modal information from visible (VIS) and thermal infrared (TIR) images is crucial for object detection in fully adapting to varied lighting conditions. However, the existing models usually treat VIS and TIR images as independent information and extract corresponding features from separate networks due to the scarcity of training data with labeled instances from both VIS and TIR registration images. To fill this gap, a novel Multi-Modal Fusion NETwork (M2FNet) based on the Transformer architecture is proposed in this paper, which contains two effective modules: the Union-Modal Attention (UMA) and the Cross-Modal Attention (CMA). The UMA module aggregates multi-spectral features from VIS and TIR images and then extracts multi-modal features via a convolutional neural network (CNN) backbone. The CMA module is designed to learn cross-attention features from VIS and TIR pairwise features by Transformer architecture. Evaluation results by the mean average precision (mAP) metric show that the M2FNet method significantly advances the baseline methods trained using only VIS or TIR images by 10.71 % and 2.97 %, respectively. The increments in mAP are observed in the M2FNet method compared with the existing multi-modal methods on two public datasets. Sensitivity analysis of eight illumination thresholds shows that the M2FNet method presents robustness performance on varied illumination conditions and achieves the maximum increase in accuracy of 25.6 %. Moreover, this method is subsequently applied to a new testing dataset, VI2DA (Visible-Infrared paired Video and Image DAtaset), observed by diverse sensors and platforms for testing the generalization ability of object detectors, which will be publicly available at https://github.com/TIR-OD/Datasets.

Tartronic acid is known for its potential to inhibit sugar-to-lipid conversion in the human body, leading to weight loss and fat reduction. This compound is predominantly found in cucumbers and other cucurbit crops. Therefore, cultivating cucumbers with high tartronic acid content holds significant health implications. In this study, we assessed the tartronic acid content in 52 cucumber germplasms with favorable overall traits and identified 8 cucumber germplasms with elevated tartronic acid levels. Our investigation into factors influencing cucumber tartronic acid revealed a decrease in content with fruit development from the day of flowering. Furthermore, tartronic acid content was higher in early-harvested fruits compared to late-harvested ones, with the rear part of the fruit exhibiting significantly higher content than other parts. Foliar spraying of microbial agents increased tartronic acid content by 84.4%. This study provides valuable resources for breeding high tartronic acid cucumbers and offers practical insights for optimizing cucumber production practices.

Hyperspectral remote sensing technology is an advanced and powerful tool that enables fine identification of the numerous soil reflectance spectrum characteristics. Heavy metal(loid)s (HMs) are the primary pollutants affecting the soil biodiversity and ecosystem services. Estimating HMs’ concentrations in soils using hyperspectral data is an effective method but is challenging due to the effects of varied soil properties and measurement-related errors inflicted by atmospheric effects. This study focused on typical mining areas in the Fenwei Plain (FWP), China. Soil-related data were collected by leveraging airborne- and ground-based integrated remote sensing observations. The concentrations of eight HMs, namely copper (Cu), lead (Pb), zinc (Zn), nickel (Ni), chromium (Cr), cadmium (Cd), arsenic (As), and mercury (Hg), were measured by laboratory analysis from 100 in situ soil samples. Soil reflectance spectra were processed based on resampling and envelope methods. The combination datasets of the concentrations and optimal soil reflectance spectra were used to build the soil-related parameter retrieval models using three machine learning (ML) methods, and the feasibility of applying the high-performance retrieval model to estimate the HM concentrations in mining areas was evaluated and explored. Spectral analysis results show that four hundred and twenty-eight bands of five wavelength ranges are of high quality and obviously demonstrate the spectral characteristics selected to build the soil-related parameter models. The evaluation results of eight combination data subsets and three methods show that the preprocessing of spectral data (ground- and airborne-based reflectance) and soil samples with the random forest (RF) method can obtain higher accuracy than support vector machine (SVM) and partial least squares (PLS) methods, denoted as the AER-ACS-RF and GER-GCS-RF models (the average RMSE values of eight HMs were 2.61 and 2.53 mg/kg, respectively). The highest R2 values were observed in Cd and As, with an equal value of 0.98, followed by that of Pb (R2 = 0.97). The relative prediction deviation (RPD) values of Cu and AS were greater than 1.9. Moreover, the airborne-based AER-ACS-RF model presents a good mapping effect about the concentrations (mg/kg) of eight HMs in mining areas, ranging from 21.65 to 31.25 (Cu), 16.38 to 30.45 (Pb), 62.02 to 109.48 (Zn), 23.33 to 32.47 (Ni), 49.81 to 66.56 (Cr), 0.09 to 0.23 (Cd), 7.31 to 12.24 (As), and 0.03 to 0.17 (Hg), respectively.

Land surface temperature (LST) is an important parameter affecting ground-air energy exchange, and can be applied in many fields such as drought monitoring, evapotranspiration estimation, and crop yield assessment. The single-channel (SC) algorithm only needs thermal infrared (TIR) data with only one channel to retrieve the LST, but requires accurate atmospheric correction, which is difficult to guarantee in many cases. The split-window (SW) algorithm exploits the difference in atmospheric absorption of adjacent channels and can accurately obtain the LST without atmospheric parameters, but the input requirement of two TIR channels makes the application limited. Some previous studies found that there is a strong correlation between the top-of-atmosphere thermal radiance in adjacient TIR channels with similar wavelength ranges, which makes it possible to decompose a single TIR channel into two TIR channels with close wavelengths using empirical relationships with errors smaller than those caused by the uncertainty of atmospheric correction. In this article, Landsat-7 ETM+ TIR data was decomposed into two virtual TIR channels by fusing with the simultaneously observed Terra moderate resolution imaging spectroradiometer (MODIS) data, and the channel decomposed SW (TCD-SW) based on the generalized nonlinear SW algorithm was developed and applied to retrieve the LST to eliminate the dependence on atmospheric parameters. The validation results using the simulation dataset, ground-measured site data, and well-validated MODIS sea surface temperature product showed that the proposed TCD-SW algorithm achieved more accurate results than the SC algorithm, being more advantageous in the humid atmosphere. The TCD-SW algorithm can be used as a potential new method for LST retrieval from remote sensing data with only one TIR channel under complex atmospheric conditions.

Land surface temperature (LST) is one of important parameter in the earth energy and water budget. Recently, open-source modules or packages are emerging as all-in-one solutions to speed up the rapid development and critical application of LST from thermal infrared (TIR) remote sensing data. In this article, we developed a python package, dubbed Peking University land surface temperature (PKULAST) model to address the retrieval of LST from TIR remote sensing data with different LST retrieval algorithms, such as single channel algorithm, split window algorithm, and the temperature and emissivity separation algorithm. extendable data structures in the model covers common conceptual models used in LST algorithms implementation, such as spectral response functions, surface spectral libraries (ASTER, ECOSTRESS etc.), atmospheric profiles (NCEP, ERA5, MERRA-2, etc.), making it a convenience to construct the processing chain of developing new LST retrieval algorithms and generating LST products. The ASTER imageries are tested to assess the performance of the PKULAST to derive accurate LST. The LST result is then evaluated using long-term ground-based longwave radiation observations collected at seven sites from 2000 to 2021 under cloud-free scenes, and the result shows that the PKULAST model can obtain LST with an uncertainty of 2.7 K in bias, and 4.4 K in root-mean-square error. With the help of predefined extendable data structures and workflows for different sensors, atmospheric profile libraries and surface spectral libraries, new algorithms for retrieving LST can be developed easily. The PKULAST promises to be a tool that seamlessly interoperates with developing LST retrieval algorithms and generating LST scientific products. The code will be released at https://github.com/tirzhu/pkulast.

Object detection is one of the most crucial tasks in computer vision and remote sensing to identify specific categories of various objects in images. The unmanned aerial vehicle (UAV)-based thermal infrared (TIR) remote sensing multi-scenario images and videos are two important data sources in public security. However, their object detection process is still challenging because of the complicated scene information, coarse resolution compared with the visible videos and lack of public labelled datasets and training models. This study proposed a UAV TIR object detection framework for images and videos. The You Only Look Once (YOLO) models based on Convolutional Neural Network (CNN) architecture were designed to extract features from ground-based TIR images and videos, which were captured by Forward-looking Infrared (FLIR) cameras. The most effective algorithm was finally identified by evaluation metrics and then applied to detect objects on TIR videos from UAVs. Results showed that the highest mean average precision (mAP) of the person and car instances was 88.69% in the validating task. The fastest detection speed achieved 50 frames per second (FPS), and the smallest model size was observed in YOLOv5-s. In the application, the cross-detection performance on persons and cars in UAV TIR videos under a YOLOv5-s model was discussed in terms of the different UAVs’ observation angles and the effectiveness of the YOLO architecture was revealed. This study provides positive support for the qualitative and quantitative evaluation of objection detection from TIR images and videos using deep-learning models.

Accurate and timely crop yield estimation is critical for food security and sustainable development. The rapid development of unmanned aerial vehicles (UAVs) offers a new approach to acquire high spatio-temporal resolution imagery of farmland at a low cost. In order to realize the full potential of UAV platform and sensor, machine learning has been introduced to estimate crop yield, but the shortages of field measurements have troubled researchers. In this article, the CW-RF model, a new wheat yield estimation model suitable for the North China plain, was established using random forest, and the crop growth model (the CERES-wheat model) was chosen to simulate abundant training samples for random forest at field plot scale. According to CERES-wheat model simulation, the leaf area index (LAI) and leaf nitrogen content (LNC) at the wheat jointing and heading stages were selected as the most sensitive parameters, and were retrieved from UAV hyperspectral imagery using the directional second derivative and angular insensitivity vegetation index methods, respectively. Then the retrieved LAI and LNC results were input into the CW-RF model to estimate winter wheat yield. The field validation in Luohe, Henan showed that the root-mean-squared error of the retrieved LAI and LNC were 6.27% and 12.17% at jointing stages, 9.21% and 13.64% at heading stages, respectively. The RMSE of estimated yield was 1,008.08 kg/ha, and the mean absolute percent error of estimated yield was 9.36%, demonstrating the available of the CW-RF model in wheat yield estimation at field plot scale. Apart from Luohe, validations in some other fields (e.g., Xiaotangshan, Beijing), prove the wide applicability of the CW-RF model. In addition, the UAV hyperspectral data were found to significantly improve the retrieval accuracy, and further improve CW-RF model estimation accuracy. In conclusion, this article showed that the CERES-Wheat model simulation can be important data source for machine learning-based wheat yield estimation model at field plot scale, and the hyperspectral sensor mounted on a UAV is a feasible remote sensing data acquisition mode for winter wheat growth monitoring and yield estimation.

The second-generation Chinese polar-orbit meteorological satellite Fengyun-3D carries the MEdium-Resolution Spectral Imager (MERSI-II). MERSI-II has 19 solar channels and 6 middle and thermal infrared channels and can provide daily global coverage observation at daytime and nighttime. Land surface temperature (LST) products have a wide range of application requirements. Unfortunately, there is no official LST product of MERSI-II yet. This article proposed a temperature-emissivity separation algorithm to estimate LST and emissivity from two mid-infrared (MIR) and two thermal infrared (TIR) nighttime MERSI-II images. The sensitive analysis and ground validation show the following. First, the algorithm can theoretically retrieve LST with an error of less than 0.7 K, emissivity with errors of about 0.025 for MIR channel and 0.01 for TIR channel, respectively. Second, the nighttime LST retrieval error was approximately 2.19 K, and the bias was approximately 0.60 K among 6 SURFRAD sites and 2 PKULSTNet sites. The retrieved emissivity was cross-validated using MODIS emissivity product over a desert area and its difference was obtained as 0.01 and 0.016 for two TIR channels. Finally, the TES algorithm was applied to obtain LST and emissivity images over North China Plain as an example, and the nighttime LST and land surface emissivity (LSE) were obtained well. As a result, this study shows that the proposed TES algorithm provides an effective way to get LST and LSE image from FY-3D/ MERSI-II nighttime observation, which will improve their applications in different fields.

Land surface temperature (LST) retrieval from thermal infrared (TIR) remote sensing image requires atmospheric and land surface emissivity (LSE) data that are sometimes unattainable. To overcome this problem, a hybrid algorithm is developed to retrieve LST without atmospheric correction and LSE data input, by combining the split-window (SW) and temperature–emissivity separation (TES) algorithms. The SW algorithm is used to estimate surface-emitting radiance in adjacent TIR bands, and such radiance is applied to the TES algorithm to retrieve LST and LSE. The hybrid algorithm is implemented on five TIR bands of the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER). Analysis shows that the hybrid algorithm can estimate LST and LSE with an error of 0.5–1.5 K and 0.007–0.020, respectively. Moreover, the LST error of the hybrid algorithm is equivalent to that of the original ASTER TES algorithm, involving 1%–2% uncertainty in atmospheric correction. The hybrid algorithm is validated using ground-measured LST at six sites and ASTER LST products, indicating that the temperature difference between the ASTER TES algorithm and the hybrid algorithm is 1.4 K and about 2.5–3.5 K compared to the ground measurement. Finally, the hybrid algorithm is applied to at two places.

Land surface component temperatures (LSCTs), i.e., the temperatures of soil and vegetation, are important parameters in many applications, such as estimating evapotranspiration and monitoring droughts. However, the multiangle algorithm is affected due to different spatial resolution between nadir and oblique views. Therefore, we propose a combined retrieval algorithm that uses dual-angle and multipixel observations together. The sea and land surface temperature radiometer onboard ESA's Sentinel-3 satellite allows for quasi-synchronous dual-angle observations, from which LSCTs can be retrieved using dual-angle and multipixel algorithms. The better performance of the combined algorithm is demonstrated using a sensitivity analysis based on a synthetic dataset. The spatial errors in the oblique view due to different spatial resolution can reach 4.5 K and have a large effect on the multiangle algorithm. The introduction of multipixel information in a window can reduce the effect of such spatial errors, and the retrieval results of LSCTs can be further improved by using multiangle information for a pixel. In the validation, the proposed combined algorithm performed better, with LSCT root mean squared errors of 3.09 K and 1.91 K for soil and vegetation at a grass site, respectively, and corresponding values of 3.71 K and 3.42 K at a sparse forest site, respectively. Considering that the temperature differences between components can reach 20 K, the results confirm that, in addition to a pixel-average LST, the combined retrieval algorithm can provide information on LSCTs. This article demonstrates the potential of utilizing additional information sources for better LSCT results, which makes the presented combined strategy a promising option for deriving large-scale LSCT products.

Rhizome cutting is prevalent in propagation for Hemerocallis citrina Baroni. This study aimed to reveal the influence of rhizome weight on plantlet growth, and evaluate a new cutting protocol. Three levels of rhizome weight (big (BR), medium (MR) and small (SR)) were compared by measuring plantlet growth four times during cutting propagation. Sectioning rhizomes longitudinally along shaft axis of crown into 2, 3 or 4 parts (S2, S3 and S4), was tested on rhizomes with two bud sizes (Large bud (LB), Small bud (SB)). Propagation coefficient, the number of plantlets obtained per rhizome, kept similar regardless of rhizome weight, while S3 and S4 increased propagation coefficient. Rhizome weight changed the relative growth rates on different dates. SR resulted in lower aboveground dry biomass, leaf area, total fibrous root length, relative total fibrous root length growth rate and N deficiency compared to BR. Sucrose concentration in rhizome decreased with time and fructose concentration was lowest in MR at the end of experiment. Aboveground dry biomass and total fibrous root length were lowest in S4, followed by S3 and S2. Rhizome weight and number of sectioning per rhizome influence cutting propagation, and sectioning rhizomes into three pieces has higher propagation coefficient and less retardation on growth. According to N deficiency in plantlets during the propagation, N fertiliser is probably needed during cutting.

Thermal infrared (TIR) remote sensing technology is capable of acquiring large-scale land surface temperature (LST), which is a key factor in the energy exchange between land surface and atmosphere. The visible and infrared multispectral sensor (VIMS) equipped in the Chinese Gaofen-5 (GF-5) satellite can obtain four channels of TIR images with a 40 m spatial resolution. However, due to the change of working environment, the TIR sensor suffers a low-accurate radiometric calibration that needs improvement. This paper puts forward a new cross-calibration for GF-5/VIMS TIR images by linking the top of the atmosphere (TOA) radiance of the vertical angle MODIS observation with the GF-5/VIMS image to estimate the radiometric calibration coefficients, Gain and Offset. To verify the recalibration performance, a new nonlinear two-channel split-window (SW) algorithm and a light gradient boosting machine (LightGBM) method which is used to refine LST from the SW algorithm by minimizing the residuals, were developed for the recalibrated GF-5/VIMS TIR 3 and TIR 4 channels. The radiometric cross-calibration algorithm and the optimized SW algorithm were applied to real GF-5/VIMS TIR images. The validation results showed that the brightness temperature and LST were improved significantly, and the LST retrieval error was reduced greatly to 1.79 K after recalibration, indicating a large improvement of the LST retrieval for GF-5/VIMS TIR image using the proposed cross-calibration and SW algorithms.

Abstract The lunar surface temperature (LST) derived from thermal infrared (TIR) measurements can aid in understanding the physical properties of the lunar surface. The Diviner Lunar Radiometer Experiment (herein, Diviner) sensor provides global lunar surface observation in seven TIR channels. However, its retrieval of LST constantly uses a single emissivity value (i.e., 0.95) by ignoring the spatial variation of lunar surface, thereby reducing the accuracy of temperature and day–night temperature difference. To overcome this problem, this study developed a physical method called temperature–emissivity separation (TES) algorithm to retrieve LST and lunar surface emissivity from the daytime observation in three Christiansen Feature (CF) channels (7.55–8.05, 8.10–8.40, and 8.38–8.60 μm) of the Diviner, and then used the emissivity from daytime observation to inverse LST at nighttime observation. Findings showed that the TES algorithm could retrieve LST and emissivity with an error of less than 0.8 K and 0.008, respectively. However, observation noise significantly affected the retrieval accuracy, particularly for the low‐temperature pixels; moreover, high retrieval accuracy requires a surface temperature higher than 240 K. The new algorithm was applied to obtain the daytime and nighttime LST and emissivity from the Diviner images. Results showed that the LST retrieved from the algorithm differed approximately 3.9 K from that calculated from a single emissivity 0.95. Finally, an example of global surface temperature and emissivity were obtained. Consequently, the CF pixels were found to distribute in the latitude range from −60° to 60°; however, they did not have a large distribution in high‐latitude and near‐polar regions.

Cold tolerance is improved by cold stress acclimation (CS-ACC), and the cold tolerance level is ‘remembered’ by plants. However, the underlying signaling mechanisms remain largely unknown. Here, the CS memory mechanism was studied by bioinformation, plant physiological and photosynthetic parameters, and gene expression. We found that CS-ACC induced the acquisition of CS memory and enhanced the maintenance of acquired cold tolerance (MACT) in cucumber seedlings. The H2O2 content and NADPH oxidase activity encoded by CsRBOH was maintained at higher levels during recovery after CS-ACC and inhibition of RBOH-dependent signaling after CS-ACC resulted in a decrease in the H2O2 content, NADPH oxidase activity, and MACT. CsRBOH2, 3, 4, and 5 showed high expression during recovery after CS-ACC. Many BZR-binding sites were identified in memory-responsive CsRBOHs promoters, and CsBZR1 and 3 showed high expression during recovery after CS-ACC. Inhibition of RBOH-dependent signaling or brassinosteroids affected the maintenance of the expression of these memory-responsive CsRBOHs and CsBZRs. The photosynthetic efficiency (PE) decreased but then increased with the prolonged recovery after CS-ACC, and was higher than the control at 48 h of recovery; however, inhibition of RBOH-dependent signaling resulted in a lower PE. Further etiolated seedlings experiments showed that a photosynthetic capacity was necessary for CS memory. Therefore, photosynthesis mediated by RBOH-dependent signaling is essential for CS memory.

Cucumber cultivars: Genes could offer a spiny solution A group of genes connected with spine growth in cucumber fruits could enable the development of cultivars for specific markets. In China, people prefer cucumbers with plentiful spines, while for the European market, cucumbers are usually smooth. This has implications for growers, yet the molecular mechanisms inherent in spine or 'trichome' growth are not completely clear. Yiqun Weng at the University of Wisconsin-Madison in the United States, together with Huazhong Ren at the China Agricultural University in Beijing, and co-workers identified genes related to trichomes in cucumbers. In a genome-wide survey, the researchers identified 91 genes from a family of transcription factors predicted to influence trichome growth. Of these, six genes were highly expressed in wild-type spiny cucumbers, while a further three were found highly expressed in mutant smooth cucumbers, indicating that they may reduce spine growth.

By effectively observing the land surface and obtaining farmland conditions, satellite remote sensing has played an essential role in agricultural drought monitoring over past decades. Among all remote sensing techniques, optical and thermal remote sensing have the most extended history of being utilized in drought monitoring. The primary goal of this paper is to illustrate how optical and thermal remote sensing have been and will be applied in the monitoring, assessment, and prediction of agricultural drought. We group the methods into four categories: optical, thermal, optical and thermal, and multi-source. For each category, a concise explanation is given to show the inherent mechanisms. We pay special attention to solar-induced chlorophyll fluorescence, which has great potential in early drought detection. Finally, we look at the future directions of agricultural drought monitoring, including (1) early detection; (2) spatio-temporal resolution; (3) organic combination of multi-source data; and (4) smart prediction and assessment based on deep learning and cloud computing.

Gross primary productivity (GPP) represents total vegetation productivity and is crucial in regional or global carbon balance. The Northeast China (NEC), abundant in vegetation resources, has a relatively large vegetation productivity; however, under obvious climate change (especially warming), whether and how will the vegetation productivity and ecosystem function of this region changed in a long time period needs to be revealed. With the help of GPP products provided by the Global LAnd Surface Satellite (GLASS) program, this paper gives an overview of the regional feedback of vegetation productivity to the changing climate (including temperature, precipitation, and solar radiation) across the NEC from 1982 to 2015. Analyzing results show a slight positive response of vegetation productivities to warming across the NEC with an overall increasing trend of GPPGS (accumulated GPP within the growing season of each year) at 4.95 g C/m2. yr−2 over the last three decades. More specifically, the growth of crops, rather than forests, contributes more to the total increasing productivity, which is mainly induced by the agricultural technological progress as well as warming. As for GPP in forested area in the NEC, the slight increment of GPPGS in northern, high-latitude forested region of the NEC was caused by warming, while non-significant variation of GPPGS was found in southern, low-latitude forested region. In addition, an obvious greening trend, as reported in other regions, was also found in the NEC, but GPPGS of forests in southern NEC did not have significant variations, which indicated that vegetation productivity is not bound to increase simultaneously with greening, except for these high-latitude forested areas in the NEC. The regional feedback of vegetation productivity to climate change in the NEC can be an indicator for vegetations growing in higher latitudes in the future under continued climate change.

The protection of water resources is of paramount importance to human beings’ practical lives. Monitoring and improving water quality nowadays has become an important topic. In this study, a novel Bayesian probabilistic neural network (BPNN) improved from ordinary Bayesian probability methods has been developed to quantitatively predict water quality parameters including phosphorus, nitrogen, chemical oxygen demand (COD), biochemical oxygen demand (BOD), and chlorophyll a. The proposed method, based on conventional Bayesian probability methods, involves feature engineering and deep neural networks. Additionally, it extracts significant information for each endmember from combinations of spectra by feature extraction, with spectral unmixing based on mathematical and statistical analysis, and calculates each of the water quality parameters. The experimental results show the great performance of the proposed model with all coefficient of determination R 2 over 0.9 greater than the values (0.6–0.8) from conventional methods, which are greater than ordinary Bayesian probability analysis. The mean percent of absolute error (MPAE) is taken into account as an important statistical criterion to evaluate model performance, and our results show that MPAE ranges from 4% (nitrogen) to 10% (COD). The root mean squared errors (RMSEs) of phosphorus, nitrogen, COD, BOD, and chlorophyll-a (Chla) are 0.03 mg/L, 0.28 mg/L, 3.28 mg/L, 0.49 mg/L, and 0.75 μg/L, respectively. In comparison with other deep learning methods, this study takes a relatively small amount of data as training data to train the proposed model and the proposed model is then tested on the same amount of testing data, achieving a greater performance. Thus, the proposed method is time-saving and more effective. This study proposes a more compatible and effective method to assist with decomposing combinations of hyperspectral signatures in order to calculate the content level of each water quality parameter. Moreover, the proposed method is practically applied to hyperspectral image data on board an unmanned aerial vehicle in order to monitor the water quality on a large scale and trace the location of pollution sources in the Maozhou River, Guangdong Province of China, obtaining well-explained and significant results.

Landsat 8/thermal infrared sensor (TIRS) is suffering from the problem of stray light that makes an inaccurate radiance for two thermal infrared (TIR) bands and the latest correction was conducted in 2017. This paper focused on evaluation of land surface temperature (LST) retrieval from Landsat 8 before and after the correction using ground-measured LST from five surface radiation budget network (SURFRAD) sites. Results indicated that the correction increased the band radiance at the top of the atmosphere for low temperature but decreased such radiance for high temperature. The root-mean-square error (RMSE) of LST retrieval decreased by 0.27 K for Band 10 and 0.78 K for Band 11 using the single-channel algorithm. For the site with high temperature, the LST retrieval RMSE of the single-channel algorithm for Band 11 even reduced by 1.4 K. However, the accuracy of two of three split-window algorithms adopted in this paper decreased. After correction, the single-channel algorithm for Band 10 and the linear split-window algorithm had the least RMSE (approximately 2.5 K) among five adopted algorithms. Moreover, besides SURFRAD sites, it is necessary to validate using more robust and homogeneous ground-measured datasets to help solely clarify the effect of the correction on LST retrieval.

Protection of water environments is an important part of overall environmental protection; hence, many people devote their efforts to monitoring and improving water quality. In this study, a self-adapting selection method of multiple artificial neural networks (ANNs) using hyperspectral remote sensing and ground-measured water quality data is proposed to quantitatively predict water quality parameters, including phosphorus, nitrogen, biochemical oxygen demand (BOD), chemical oxygen demand (COD), and chlorophyll a. Seventy-nine ground measured data samples are used as training data in the establishment of the proposed model, and 30 samples are used as testing data. The proposed method based on traditional ANNs of numerical prediction involves feature selection of bands, self-adapting selection based on multiple selection criteria, stepwise backtracking, and combined weighted correlation. Water quality parameters are estimated with coefficient of determination R 2 ranging from 0.93 (phosphorus) to 0.98 (nitrogen), which is higher than the value (0.7 to 0.8) obtained by traditional ANNs. MPAE (mean percent of absolute error) values ranging from 5% to 11% are used rather than root mean square error to evaluate the predicting precision of the proposed model because the magnitude of each water quality parameter considerably differs, thereby providing reasonable and interpretable results. Compared with other ANNs with backpropagation, this study proposes an auto-adapting method assisted by the above-mentioned methods to select the best model with all settings, such as the number of hidden layers, number of neurons in each hidden layer, choice of optimizer, and activation function. Different settings for ANNS with backpropagation are important to improve precision and compatibility for different data. Furthermore, the proposed method is applied to hyperspectral remote sensing images collected using an unmanned aerial vehicle for monitoring the water quality in the Shiqi River, Zhongshan City, Guangdong Province, China. Obtained results indicate the locations of pollution sources.

Multi-angular observation of land surface thermal radiation is considered to be a promising method of performing the angular normalization of land surface temperature (LST) retrieved from remote sensing data. This paper focuses on an investigation of the minimum requirements of viewing angles to perform such normalizations on LST. The normally kernel-driven bi-directional reflectance distribution function (BRDF) is first extended to the thermal infrared (TIR) domain as TIR-BRDF model, and its uncertainty is shown to be less than 0.3 K when used to fit the hemispheric directional thermal radiation. A local optimum three-angle combination is found and verified using the TIR-BRDF model based on two patterns: the single-point pattern and the linear-array pattern. The TIR-BRDF is applied to an airborne multi-angular dataset to retrieve LST at nadir (Te-nadir) from different viewing directions, and the results show that this model can obtain reliable Te-nadir from 3 to 4 directional observations with large angle intervals, thus corresponding to large temperature angular variations. The Te-nadir is generally larger than temperature of the slant direction, with a difference of approximately 0.5~2.0 K for vegetated pixels and up to several Kelvins for non-vegetated pixels. The findings of this paper will facilitate the future development of multi-angular thermal infrared sensors.

Fractional vegetation cover (FVC) is one of the most important criteria for surface vegetation status. This criterion corresponds to the complement of gap fraction unity at the nadir direction and accounts for the amount of horizontal vegetation distribution. This study aims to directly validate the accuracy of FVC products over crops at coarse resolutions (1 km) by employing field measurements and high-resolution data. The study area was within an oasis in the Heihe Basin, Northwest China, where the Heihe Watershed Allied Telemetry Experimental Research was conducted. Reference FVC was generated through upscaling, which fitted field-measured data with spaceborne and airborne data to retrieve high-resolution FVC, and then high-resolution FVC was aggregated with a coarse scale. The fraction of green vegetation cover product (i.e., GEOV1 FVC) of SPOT/VEGETATION data taken during the GEOLAND2 project was compared with reference data. GEOV1 FVC was generally overestimated for crops in the study area compared with our estimates. Reference FVC exhibits a systematic uncertainty, and GEOV1 can overestimate FVC by up to 0.20. This finding indicates the necessity of reanalyzing and improving GEOV1 FVC over croplands.

This paper developed a practical split-window (SW) algorithm to estimate land surface temperature (LST) from Thermal Infrared Sensor (TIRS) aboard Landsat 8. The coefficients of the SW algorithm were determined based on atmospheric water vapor sub-ranges, which were obtained through a modified split-window covariance–variance ratio method. The channel emissivities were acquired from newly released global land cover products at 30 m and from a fraction of the vegetation cover calculated from visible and near-infrared images aboard Landsat 8. Simulation results showed that the new algorithm can obtain LST with an accuracy of better than 1.0 K. The model consistency to the noise of the brightness temperature, emissivity and water vapor was conducted, which indicated the robustness of the new algorithm in LST retrieval. Furthermore, based on comparisons, the new algorithm performed better than the existing algorithms in retrieving LST from TIRS data. Finally, the SW algorithm was proven to be reliable through application in different regions. To further confirm the credibility of the SW algorithm, the LST will be validated in the future.

The radiometric performance of remotely-sensed images is important for the applications of such data in monitoring land surface, ocean and atmospheric status. One requirement placed on the Thermal Infrared Sensor (TIRS) onboard Landsat 8 was that the noise-equivalent change in temperature (NEΔT) should be ≤0.4 K at 300 K for its two thermal infrared bands. In order to optimize the use of TIRS data, this study investigated the on-orbit NEΔT of the TIRS two bands from a scene-based method using clear-sky images over uniform ground surfaces, including lake, deep ocean, snow, desert and Gobi, as well as dense vegetation. Results showed that the NEΔTs of the two bands were 0.051 and 0.06 K at 300 K, which exceeded the design specification by an order of magnitude. The effect of NEΔT on the land surface temperature (LST) retrieval using a split window algorithm was discussed, and the estimated NEΔT could contribute only 3.5% to the final LST error in theory, whereas the required NEΔT could contribute up to 26.4%. Low NEΔT could improve the application of TIRS images. However, efforts are needed in the future to remove the effects of unwanted stray light that appears in the current TIRS images.

Studying wheat phenology can greatly enhance our understanding of how wheat growth responds to climate change, and guide us to reasonably confront its influence. However, comprehensive global-scale wheat phenology–climate analysis is still lacking. In this study, we extracted the wheat harvest date (WHD) from 1981–2014 from satellite data using threshold-, logistic-, and shape-based methods. Then, we analyzed the effects of heat and drought stress on WHD based on gridded daily temperature and monthly drought data (the Palmer drought severity index (PDSI) and the standardized precipitation evapotranspiration index (SPEI)) over global wheat-growing areas. The results show that WHD was generally delayed from the low to mid latitudes. With respect to variation trends, we detected a significant advancement of WHD in 32.1% of the world’s wheat-growing areas since 1981, with an average changing rate of −0.25 days/yr. A significant negative correlation was identified between WHD and the prior three months’ normal-growing-degree-days across 50.4% of the study region, which implies that greater preseason effective temperature accumulation may cause WHD to occur earlier. Meanwhile, WHD was also found to be significantly and negatively correlated with the prior three months’ extreme-growing-degree-days across only 9.6% of the study region (mainly located in northern South Asia and north Central-West Asia). The effects of extreme heat stress were weaker than those of normal thermal conditions. When extreme drought (measured by PDSI/SPEI) occurred in the current month, in the month prior to WHD, and in the second month prior to WHD, it forced WHD to advance by about 9.0/8.1 days, 13.8/12.2 days, and 10.8/5.3 days compared to normal conditions, respectively. In conclusion, we highlight the effects that heat and drought stress have on advancing wheat harvest timing, which should be a research focus under future climate change.

Land surface temperature (LST) is a crucial parameter in the interaction between the ground and the atmosphere. The Sentinel-3A Sea and Land Surface Temperature Radiometer (SLSTR) provides global daily coverage of day and night observation in the wavelength range of 0.55 to 12.0 μm. LST retrieved from SLSTR is expected to be widely used in different fields of earth surface monitoring. This study aimed to develop a split-window (SW) algorithm to estimate LST from two-channel thermal infrared (TIR) and one-channel middle infrared (MIR) images of SLSTR observation. On the basis of the conventional SW algorithm, using two TIR channels for the daytime observation, the MIR data, with a higher atmospheric transmittance and a lower sensitivity to land surface emissivity, were further used to develop a modified SW algorithm for the nighttime observation. To improve the retrieval accuracy, the algorithm coefficients were obtained in different subranges, according to the view zenith angle, column water vapor, and brightness temperature. The proposed algorithm can theoretically estimate LST with an error lower than 1 K on average. The algorithm was applied to northern China and southern UK, and the retrieved LST captured the surface features for both daytime and nighttime. Finally, ground validation was conducted over seven sites (four in the USA and three in China). Results showed that LST could be estimated with an error mostly within 1.5 to 2.5 K from the algorithm, and the error of the nighttime algorithm involved with MIR data was about 0.5 K lower than the daytime algorithm.

Nitric oxide (NO) is a gaseous signaling molecule in plants, transducing information as a result of exposure to low temperatures. However, the underlying molecular mechanism linking NO with chilling stress is not well understood. Here, we functionally characterized the cucumber (Cucumis sativus) nitric oxide synthase-associated gene, NITRIC OXIDE ASSOCIATED 1(CsNOA1). Expression analysis of CsNOA1, using quantitative real-time PCR, in situ hybridization and a promoter::β-glucuronidase (GUS) reporter assay, revealed that it is expressed mainly in the root and shoot apical meristem (SAM), and that expression is up-regulated by low temperatures. A CsNOA1-GFP fusion protein was found to be localized to the mitochondria, and ectopic expression of CsNOA1 in the A. thaliana noa1 mutant partially rescued the normal phenotype. Transgenic cucumber plants revealed that the gene is required by seedlings to tolerate chilling stress: constitutive over-expression of CsNOA1 led to a greater accumulation of soluble sugars, starch, and an up-regulation of Cold-regulatory C-repeat binding factor3 (CBF3) expression. Conversely, suppression of CsNOA1 expression resulted in the opposite phenotype and a reduced NO content compared to wild type plants. Additionally, analyses using inhibitors suggested that NO can be synthesized via a CsNOA1 dependent pathway that is independent of the nitrate reductase (NR) pathway.

Cucumber (Cucumis sativus L.) is threatened by substantial yield losses due to the south root-knot nematode (Meloidogyne incognita). However, understanding of the molecular mechanisms underlying the process of nematode infection is still limited. In this study, we found that M. incognita infection affected the structure of cells in cucumber roots and treatment of the cytoskeleton inhibitor (cytochalasin D) reduced root-knot nematode (RKN) parasitism. It is known that Actin-Depolymerizing Factor (ADF) affects cell structure, as well as the organization of the cytoskeleton. To address the hypothesis that nematode-induced abnormal cell structures and cytoskeletal rearrangements might be mediated by the ADF genes, we identified and characterized eight cucumber ADF (CsADF) genes. Phylogenetic analysis showed that the cucumber ADF gene family is grouped into four ancient subclasses. Expression analysis revealed that CsADF1, CsADF2-1, CsADF2-2, CsADF2-3 (Subclass I) and CsADF6 (Subclass III) have higher transcript levels than CsADF7-1, CsADF7-2 (Subclass II genes) and CsADF5 (Subclass IV) in roots. Members of subclass I genes (CsADF1, CsADF2-1, CsADF2-2 and CsADF2-3), with the exception of CsADF2-1, exhibited a induction of expression in roots 14 days after their inoculation (DAI) with nematodes. However, the expression of subclass II genes (CsADF7-1 and CsADF7-2) showed no significant change after inoculation. The transcript levels of CsADF6 (Subclass III) showed a specific induction at 21 DAI, while CsADF5 (Subclass IV) was weakly expressed in roots, but was strongly up-regulated as early as 7 DAI. In addition, treatment of roots with cytochalasin D caused an approximately two-fold down-regulation of the CsADF genes in the treated plants. These results suggest that CsADF gene mediated actin dynamics are associated with structural changes in roots as a consequence of M. incognita infection.

Trichomes are specialized epidermal cells located in aerial parts of plants that function in plant defense against biotic and abiotic stresses. The simple unicellular trichomes of Arabidopsis serve as an excellent model to study the molecular mechanism of cell differentiation and pattern formation in plants. Loss-of-function mutations in Arabidopsis thaliana have suggested that the core genes GL1 (which encodes a MYB transcription factor) and TTG1 (which encodes a WD40 repeat-containing protein) are important for the initiation and spacing of leaf trichomes, while for normal trichome initiation, the genes GL3 and EGL3 (which encode a bHLH protein) are needed. However, the positive regulatory genes involved in multicellular trichrome development in cucumber remain unclear. This review focuses on the phenotype of mutants (csgl3, tril, tbh, mict and csgl1) with disturbed trichomes in cucumber and then infers which gene(s) play key roles in trichome initiation and development in those mutants. Evidence indicates that MICT, TBH and CsGL1 are allelic with alternative splicing. CsGL3 and TRIL are allelic and override the effect of TBH, MICT and CsGL1 on the regulation of multicellular trichome development; and affect trichome initiation. CsGL3, TRIL, MICT, TBH and CsGL1 encode HD-Zip proteins with different subfamilies. Genetic and molecular analyses have revealed that CsGL3, TRIL, MICT, TBH and CsGL1 are responsible for the differentiation of epidermal cells and the development of trichomes. Based on current knowledge, a positive regulator pathway model for trichome development in cucumber was proposed and compared to a model in Arabidopsis. These data suggest that trichome development in cucumber may differ from that in Arabidopsis.

Crop growth in different periods influences the final yield. This study started from the agronomic mechanism of yield formation and aimed to extract useful spectral characteristics in different phenological phases, which could directly describe the final yield and dynamic contributions of different phases to the yield formation. Hyperspectral information of the winter wheat canopy was acquired during three important phases (jointing stage, heading stage, and grain-filling stage). An enhanced 2D correlation spectral analysis method modified by mutual information was proposed to identify the sensitive wavebands. The selected wavebands performed well with good mechanism interpretation and close correlation with important crop growth parameters and main physiological activities related to yield formation. The quantitative contribution proportions of plant growth in three phases to the final yield were estimated by determining the coefficients of partial least square models based on full spectral information. They were then used as single-phase weight factors to merge the selected wavebands. The support vector machine model based on the weighted spectral dataset performed well in yield prediction with satisfactory accuracy and robustness. This result would provide rapid and accurate guidance for agricultural production and would be valuable for the processing of hyperspectral remote sensing data.

Snow cover is an essential climate variable of the Global Climate Observing System. Gaofen-4 (GF-4) is the first Chinese geostationary satellite to obtain optical imagery with high spatial and temporal resolution, which presents unique advantages in snow cover monitoring. However, the panchromatic and multispectral sensor (PMS) onboard GF-4 lacks the shortwave infrared (SWIR) band, which is crucial for snow cover detection. To reach the potential of GF-4 PMS in snow cover monitoring, this study developed a novel method termed the restored snow index (RSI). The SWIR reflectance of snow cover is restored firstly, and then the RSI is calculated with the restored reflectance. The distribution of snow cover can be mapped with a threshold, which should be adjusted according to actual situations. The RSI was validated using two pairs of GF-4 PMS and Landsat-8 Operational Land Imager images. The validation results show that the RSI can effectively map the distribution of snow cover in these cases, and all of the classification accuracies are above 95%. Signal saturation slightly affects PMS images, but cloud contamination is an important limiting factor. Therefore, we propose that the RSI is an efficient method for monitoring snow cover from GF-4 PMS imagery without requiring the SWIR reflectance.

The rapid and accurate estimation of wheat production at a regional scale is crucial for national food security and sustainable agricultural development. This study developed a new gross primary productivity (GPP) estimation model (denoted as the [ACPM]), based on the effects of light, heat, soil moisture, and nitrogen content (N) on the light-use efficiency of winter wheat. The ACPM model used the quantic additivity of the environmental factors to improve the minimum form or multiple multiplication form in the previous model and thus characterized the joint effects of heat, soil moisture, and N on crop photosynthesis performance. The key parameters (i.e., light) were determined from the photosynthetically active radiation product of the Himawari-8 sensor and the fraction of photosynthetically active radiation product of Moderate Resolution Imaging Spectroradiometer (MODIS). The heat was determined from the land temperature products of MODIS. The soil moisture was obtained from the inversion using a visible and shortwave infrared drought index (VSDI), whereas the N stress of winter wheat was detected using the newly developed modified ratio vegetation index (MRVI), which could accurately obtain the spatiotemporal distribution of the leaf chlorophyll content of winter wheat. The ACPM and two other previous models (named the GPP1 and GPP2 models) were applied on the Himawari-8 and MODIS images in Hengshui City. The evaluation results, based on the ground measurement, indicated that the ACPM models exhibited the best estimate of dry aboveground biomass (DAM) and the wheat yield in Hengshui City, with errors of

Gross primary productivity (GPP) is an important parameter that represents the productivity of vegetation and responses to various ecological environments. The Greater Khingan Mountain (GKM) is one of the most important state-owned forest bases, and boreal forests, including the largest primeval cold-temperature bright coniferous forest in China, are widely distributed in the GKM. This study aimed to reveal spatiotemporal vegetation variations in the GKM on the basis of GPP products that were generated by the Global LAnd Surface Satellite (GLASS) program from 1982 to 2015. First, we explored the spatiotemporal distribution of vegetation across the GKM. Then we analyzed the relationships between GPP variation and driving factors, including meteorological elements, growing season length (GSL), and Fraction of Photosynthetically Active Radiation (FPAR), to investigate the dominant factor for GPP dynamics. Results demonstrated that (1) the spatial distribution of accumulated GPP (AG) in spring, summer, autumn, and the growing season varied due to three main reasons: understory vegetation, altitude, and land cover; (2) interannual AG in summer, autumn, and the growing season significantly increased at the regional scale during the past 34 years under climate warming and drying; (3) interannual changes of accumulated GPP in the growing season (AGG) at the pixel scale displayed a rapid expansion in areas with a significant increasing trend (p < 0.05) during the period of 1982–2015 and this trend was caused by the natural forest protection project launched in 1998; and finally, (4) an analysis of driving factors showed that daily sunshine duration in summer was the most important factor for GPP in the GKM and this is different from previous studies, which reported that the GSL plays a crucial role in other areas.

This paper proposes a modified model based on the PROSPECT-5 model to simulate the spectral reflectance of copper-stressed leaves. Compared with PROSPECT-5, the modified model adds the copper content of leaves as one of input variables, and the specific absorption coefficient related to copper (Kcu) was estimated and fixed in the modified model. The specific absorption coefficients of other biochemical components (chlorophyll, carotenoid, water, dry matter) were the same as those in PROSPECT-5. Firstly, based on PROSPECT-5, we estimated the leaf structure parameters (N), using biochemical contents (chlorophyll, carotenoid, water, and dry matter) and the spectra of all the copper-stressed leaves (samples). Secondly, the specific absorption coefficient related to copper (Kcu) was estimated by fitting the simulated spectra to the measured spectra using 22 samples. Thirdly, other samples were used to verify the effectiveness of the modified model. The spectra with the new model are closer to the measured spectra when compared to that with PROSPECT-5. Moreover, for all the datasets used for validation and calibration, the root mean square errors (RMSEs) from the new model are less than that from PROSPECT-5. The differences between simulated reflectance and measured reflectance at key wavelengths with the new model are nearer to zero than those with the PROSPECT-5 model. This study demonstrated that the modified model could get more accurate spectral reflectance from copper-stressed leaves when compared with PROSPECT-5, and would provide theoretical support for monitoring the vegetation stressed by copper using remote sensing.

Land surface temperature (LST) is a key variable in the study of the energy exchange between the land surface and the atmosphere. Among the different methods proposed to estimate LST, the quadratic split-window (SW) method has achieved considerable popularity. This method works well when the emissivities are high in both channels. Unfortunately, it performs poorly for low land surface emissivities (LSEs). To solve this problem, assuming that the LSE is known, the constant in the quadratic SW method was calculated by maintaining the other coefficients the same as those obtained for the black body condition. This procedure permits transfer of the emissivity effect to the constant. The result demonstrated that the constant was influenced by both atmospheric water vapour content (W) and atmospheric temperature (T0) in the bottom layer. To parameterize the constant, an exponential approximation between W and T0 was used. A LST retrieval algorithm was proposed. The error for the proposed algorithm was RMSE = 0.70 K. Sensitivity analysis results showed that under the consideration of NEΔT = 0.2 K, 20% uncertainty in W and 1% uncertainties in the channel mean emissivity and the channel emissivity difference, the RMSE was 1.29 K. Compared with AST 08 product, the proposed algorithm underestimated LST by about 0.8 K for both study areas when ASTER L1B data was used as a proxy of Gaofen-5 (GF-5) satellite data. The GF-5 satellite is scheduled to be launched in 2017.

In hermaphroditic Arabidopsis, the phytohormone gibberellin (GA) stimulates stamen development by opposing the DELLA repression of B and C classes of floral homeotic genes. GA can promote male flower formation in cucumber (Cucumis sativus L.), a typical monoecious vegetable with unisexual flowers, and the molecular mechanism remains unknown. Here we characterized a DELLA homolog CsGAIP in cucumber, and we found that CsGAIP is highly expressed in stem and male flower buds. In situ hybridization showed that CsGAIP is greatly enriched in the stamen primordia, especially during the hermaphrodite stage of flower development. Further, CsGAIP protein is located in nucleus. CsGAIP can partially rescue the plant height, stamen development and fertility phenotypes of Arabidopsis rga-24/gai-t6 mutant, and ectopic expression of CsGAIP in wide-type Arabidopsis results in reduced number of stamens and decreased transcription of B class floral homeotic genes APETALA3 (AP3) and PISTILLATA (PI). Our data suggest that monoecious CsGAIP may inhibit staminate development through transcriptional repression of B class floral homeotic genes in Arabidopsis.

The Arabidopsis SUPERMAN (SUP) gene encodes a C2H2 type zinc finger protein that is required for maintaining the boundaries between stamens and carpels, and for regulating development of ovule outer integument. Orthologs of SUP have been characterized in bisexual flowers as well as dioecious species, but it remains elusive in monoecious plants with unisexual flowers on the same individual. Here we isolate the SUP ortholog in Cucumis sativus L (CsSUP), a monoecious vegetable. CsSUP is predominantly expressed in female specific organs: the female flower buds and ovules. Ectopic expression of CsSUP in Arabidopsis can partially complement the fruit development in sup-5 mutant, and its over-expression in wide-type leads to reduced silique length, suppressed stamen development and distorted petal patterning. Our data suggest that CsSUP plays conserved as well as distinct roles during flower and fruit development, and it may function in the boundaries and ovules to balance petal patterning, stamen and ovule development in Arabidopsis.

The scaling effect correction of retrieved parameters is an essential and difficult issue in analysis and application of remote sensing information. Based on fractal theory, this paper developed a scaling transfer model to correct the scaling effect of the leaf area index (LAI) estimated from coarse spatial resolution image. As the key parameter of the proposed model, the information fractal dimension (D) of the up-scaling pixel was calculated by establishing the double logarithmic linear relationship between D-2 and the normalized difference vegetation index (NDVI) standard deviation (σNDVI) of the up-scaling pixel. Based on the calculated D and the fractal relationship between the exact LAI and the approximated LAI estimated from the coarse resolution pixel, a LAI scaling transfer model was established. Finally, the model accuracy in correcting the scaling effect was discussed. Results indicated that the D increases with increasing σNDVI, and the D-2 was highly linearly correlated with σNDVI on the double logarithmic coordinate axis. The scaling transfer model corrected the scaling effect of LAI with a maximum value of root-mean-square error (RMSE) of 0.011. The maximum absolute correction error (ACE) and relative correction error (RCE) were only 0.108% and 8.56%, respectively. The spatial heterogeneity was the primary cause resulting in the scaling effect and the key influencing factor of correction effect. The results indicated that the developed method based on fractal theory could effectively correct the scaling effect of LAI estimated from the heterogeneous pixels.