Mapping paddy rice planting areas through time series analysis of MODIS land surface temperature and vegetation index data

Modeling canopy water content in the assessment for rainfall induced surface and groundwater nitrate contamination: The Bilate cropland sub watershed

Nitrate contamination in surface and groundwater remains a widespread problem in agricultural watersheds is primarily associated to high levels of percolation or leakage from fertilized soil, which allows easy infiltration from soil into groundwater. This study was aimed to predict canopy water content to determine the nitrate contamination index resulting from nitrogen fertilizer loss in surface and groundwater. The study used Geographically Weighted Regression (GWR) model using MODIS 006 MOD13Q1-EVI Earth observation data, crop information and rainfall data. Satellite data collection was synchronized with regional crop calendars and calibrated to plant biomass. The average plant biomass during observed plant growth stages was between 0.19 kg/m2 at the minimum and 0.57 kg/m2 at the maximum. These values are based on the growth stages of crops and provide a solid basis for monitoring and validating crop water productivity data. The simulation results were validated with a high correlation coefficient (R2 = 0.996, P < 0.0005) for the observed rainfall in the growing zone compared to the predicted canopy water content. The nitrate contamination index assessment was conducted in 2004, 2008, 2009, 2010, 2011, 2013, 2014, 2015, 2018 and 2020. Canopy water content and root zone seasonal water content were measured in (%) per portion as indicators of the NO-3-N-nitrate contamination index in these years (0.391, 0.316, 0.298, 0.389, 0.380, 0.339, 0.242, 0.342 and 0.356).

A 10-m annual grazing intensity dataset in 2015-2021 for the largest temperate meadow steppe in China

Mapping grazing intensity (GI) using satellites is crucial for developing adaptive utilization strategies according to grassland conditions. Here we developed a monitoring framework based on a paired sampling strategy and the classification probability of random forest algorithm to produce annual grazing probability (GP) and GI maps at 10-m spatial resolution from 2015 to 2021 for the largest temperate meadow in China (Hulun Buir grasslands), by harmonized Landsat 7/8 and Sentinel-2 images. The GP maps used values of 0-1 to present detailed grazing gradient information. To match widely used grazing gradients, annual GI maps with ungrazed, moderately grazed, and heavily grazed levels were generated from the GP dataset with a decision tree. The GI maps for 2015-2021 had an overall accuracy of more than 0.97 having significant correlations with the statistical data at city (r = 0.51) and county (r = 0.75) scales. They also effectively captured the GI gradients at site scale (r = 0.94). Our study proposed a monitoring approach and presented annual 10-m grazing information maps for sustainable grassland management.

Mapping wetlands in Northeast China by using knowledge-based algorithms and microwave (PALSAR-2, Sentinel-1), optical (Sentinel-2, Landsat), and thermal (MODIS) images

Wetlands are rich in biodiversity, provide habitats for many wildlife species, and play a vital role in the transmission of bird-borne infectious diseases (e.g., highly pathogenic avian influenza). However, wetlands worldwide have been degraded or even disappeared due to natural and anthropogenic activities over the past two centuries. At present, major data products of wetlands have large uncertainties, low to moderate accuracies, and lack regular updates. Therefore, accurate and updated wetlands maps are needed for the sustainable management and conservation of wetlands. Here, we consider the remote sensing capability and define wetland types in terms of plant growth form (tree, shrub, grass), life cycle (perennial, annual), leaf seasonality (evergreen, deciduous), and canopy type (open, closed). We identify unique and stable features of individual wetland types and develop knowledge-based algorithms to map them in Northeast China at 10 m spatial resolution by using microwave (PALSAR-2, Sentinel-1), optical (Landsat (ETM+/OLI), Sentinel-2), and thermal (MODIS land surface temperature, LST) imagery in 2020. The resultant wetland map has a high overall accuracy of >95%. There were a total 154,254 km2 of wetlands in Northeast China in 2020, which included 27,219 km2 of seasonal open-canopy marsh, 69,158 km2 of yearlong closed-canopy marsh, and 57,878 km2 of deciduous forest swamp. Our results demonstrate the potential of knowledge-based algorithms and integrated multi-source image data for wetlands mapping and monitoring, which could provide improved data for the planning of wetland conservation and restoration.

Identification of Paddy Croplands and Its Stages Using Remote Sensors: A Systematic Review

Rice is a staple food that feeds nearly half of the world's population. With the population of our planet expected to keep growing, it is crucial to carry out accurate mapping, monitoring, and assessments since these could significantly impact food security, climate change, spatial planning, and land management. Using the PRISMA systematic review protocol, this article identified and selected 122 scientific articles (journals papers and conference proceedings) addressing different remote sensing-based methodologies to map paddy croplands, published between 2010 and October 2022. This analysis includes full coverage of the mapping of rice paddies and their various stages of crop maturity. This review paper classifies the methods based on the data source: (a) multispectral (62%), (b) multisource (20%), and (c) radar (18%). Furthermore, it analyses the impact of machine learning on those methodologies and the most common algorithms used. We found that MODIS (28%), Sentinel-2 (18%), Sentinel-1 (15%), and Landsat-8 (11%) were the most used sensors. The impact of Sentinel-1 on multisource solutions is also increasing due to the potential of backscatter information to determine textures in different stages and decrease cloud cover constraints. The preferred solutions include phenology algorithms via the use of vegetation indices, setting thresholds, or applying machine learning algorithms to classify images. In terms of machine learning algorithms, random forest is the most used (17 times), followed by support vector machine (12 times) and isodata (7 times). With the continuous development of technology and computing, it is expected that solutions such as multisource solutions will emerge more frequently and cover larger areas in different locations and at a higher resolution. In addition, the continuous improvement of cloud detection algorithms will positively impact multispectral solutions.

Coarse spatial resolution remote sensing data with AVHRR and MODIS miss the greening area compared with the Landsat data in Chinese drylands

The warming-wetting climates in Chinese drylands, together with a series of ecological engineering projects, had caused apparent changes to vegetation therein. Regarding the vegetation greening trend, different remote sensing data had yielded distinct findings. It was critical to evaluate vegetation dynamics in Chinese drylands using a series of remote sensing data. By comparing the three most commonly used remote sensing datasets [i.e., MODIS, Advanced Very High Resolution Radiometer (AVHRR), and Landsat], this study comprehensively investigated vegetation dynamics for Chinse drylands. All three remote sensing datasets exhibited evident vegetation greening trends from 2000 to 2020 in Chinese drylands, especially in the Loess Plateau and Northeast China. However, Landsat identified the largest greening areas (89.8%), while AVHRR identified the smallest greening area (58%). The vegetation greening areas identified by Landsat comprise more small patches than those identified by MODIS and AVHRR. The MODIS data exhibited a higher consistency with Landsat than with AVHRR in terms of detecting vegetation greening areas. The three datasets exhibited high consistency in identifying vegetation greening in Northeast China, Loess Plateau, and Xinjiang. The percentage of inconsistent areas among the three datasets was 39.56%. The vegetation greening areas identified by Landsat comprised more small patches. Sensors and the atmospheric effect are the two main reasons responsible for the different outputs from each NDVI product. Ecological engineering projects had a great promotion effect on vegetation greening, which can be detected by the three NDVI datasets in Chinese drylands, thereby combating desertification and reducing dust storms.

Mapping Paddy Rice in Complex Landscapes with Landsat Time Series Data and Superpixel-Based Deep Learning Method

The spatial pattern and temporal variation in paddy rice areas captured by remote sensing imagery provide an effective way of performing crop management and developing suitable agricultural policies. However, fragmented and scattered rice paddies due to undulating and varied topography, and the availability and quality of remote sensing images (e.g., frequent cloud coverage) pose significant challenges to accurate long-term rice mapping, especially for traditional pixel and phenological methods in subtropical monsoon regions. This study proposed a superpixel and deep-learning-based time series method to analyze Landsat time series data for paddy rice classification in complex landscape regions. First, a superpixel segmentation map was generated using a dynamic-time-warping-based simple non-iterative clustering algorithm with preprocessed spectral indices (SIs) time series data. Second, the SI images were overlaid onto the superpixel map to construct mean SIs time series for each superpixel. Third, a multivariate long short-term memory full convolution neural network (MLSTM-FCN) classifier was employed to learn time series features of rice paddies to produce accurate paddy rice maps. The method was evaluated using Landsat imagery from 2000 to 2020 in Cengong County, Guizhou Province, China. Results indicate that the superpixel MLSTM-FCN achieved a high performance with an overall accuracy varying from 0.9547 to 0.9721, which presents an 0.17–1.23% improvement compared to the random forest method. This study showed that combining spectral, spatial, and temporal features with deep learning methods can generate accurate paddy rice maps in complex landscape regions.

Object-Based Automatic Mapping of Winter Wheat Based on Temporal Phenology Patterns Derived from Multitemporal Sentinel-1 and Sentinel-2 Imagery

Although winter wheat has been mapped by remote sensing in several studies, such mapping efforts did not sufficiently utilize contextual information to reduce the noise and still depended heavily on optical imagery and exhausting classification approaches. Furthermore, the influence of similarity measures on winter wheat identification remains unclear. To overcome these limitations, this study developed an object-based automatic approach to map winter wheat using multitemporal Sentinel-1 (S1) and Sentinel-2 (S2) imagery. First, after S1 and S2 images were preprocessed, the Simple Non-Iterative Clustering (SNIC) algorithm was used to conduct image segmentation to obtain homogeneous spatial objects with a fusion of S1 and S2 bands. Second, the temporal phenology patterns (TPP) of winter wheat and other typical land covers were derived from object-level S1 and S2 imagery based on the collected ground truth samples, and two improved distance measures (i.e., a composite of Euclidean distance and Spectral Angle Distance, (ESD) and the difference–similarity factor distance (DSF)) were built to evaluate the similarity between two TPPs. Third, winter wheat objects were automatically identified from the segmented spatial objects by the maximum between-class variance method (OTSU) with distance measures based on the unique TPP of winter wheat. According to ground truth data, the DSF measure was superior to other distance measures in winter wheat mapping, since it achieved the best overall accuracy (OA), best kappa coefficient (Kappa) and more spatial details for each feasible band (i.e., NDVI, VV, and VH/VV), or it obtained results comparable to those for the best one (e.g., NDVI + VV). The resultant winter wheat maps derived from the NDVI band with the DSF measure achieved the best accuracy and more details, and had an average OA and Kappa of 92% and 84%, respectively. The VV polarization with the DSF measure produced the second best winter wheat maps with an average OA and Kappa of 91% and 80%, respectively. The results indicate the great potential of the proposed object-based approach for automatic winter wheat mapping for both optical and Synthetic Aperture Radar (SAR) imagery.

Research on remote sensing classification of fruit trees based on Sentinel-2 multi-temporal imageries

Accurately obtaining the spatial distribution information of fruit tree planting is of great significance to the development of fruit tree growth monitoring, disease and pest control, and yield estimation. In this study, the Sentenel-2 multispectral remote sensing imageries of different months during the growth period of the fruit trees were used as the data source, and single month vegetation indices, accumulated monthly vegetation indices (∑VIs), and difference vegetation indices between adjacent months (∆VIs) were constructed as input variables. Four conventional vegetation indices of NDVI, PSRI, GNDVI, and RVI and four improved vegetation indices of NDVIre1, NDVIre2, NDVIre3, and NDVIre4 based on the red-edge band were selected to construct a decision tree classification model combined with machine learning technology. Through the analysis of vegetation indices under different treatments and different months, combined with the attribute of Feature_importances_, the vegetation indices of different periods with high contribution were selected as input features, and the Max_depth values of the decision tree model were determined by the hyperparameter learning curve. The results have shown that when the Max_depth value of the decision tree model of the vegetation indices under the three treatments was 6, 8, and 8, the model classification was the best. The accuracy of the three vegetation index processing models on the training set were 0.8936, 0.9153, and 0.8887, and the accuracy on the test set were 0.8355, 0.7611, and 0.7940, respectively. This method could be applied to remote sensing classification of fruit trees in a large area, and could provide effective technical means for monitoring fruit tree planting areas with medium and high resolution remote sensing imageries.

Evaluating Effects of Medium-Resolution Optical Data Availability on Phenology-Based Rice Mapping in China

The phenology-based approach has proven effective for paddy rice mapping due to the unique flooding and transplanting features of rice during the early growing season. However, the method may be greatly affected if no valid observations are available during the flooding and rice transplanting phase. Here, we compare the effects of data availability of different sensors in the critical phenology phase, thereby supporting paddy rice mapping based on phenology-based approaches. Importantly, our study further analyzed the effects of the spatial pattern of the valid observations related to certain factors (i.e., sideslips, clouds, and temporal window lengths of flooding and rice transplanting), which supply the applicable area of the phenology-based approach indications. We first determined the flooding and rice transplanting phase using in situ observational data from agrometeorological stations and remote sensing data, then evaluated the effects of data availability in this phase of 2020 in China using all Landsat-7 and 8 and Sentinel-2 data. The results show that on the country level, the number of average valid observations during the flooding and rice transplanting phase was more than ten for the integration of Landsat and Sentinel images. On the sub-country level, the number of average valid observations was high in the cold temperate zone (17.4 observations), while it was relatively lower in southern China (6.4 observations), especially in Yunnan–Guizhou Plateau, which only had three valid observations on average. Based on the multicollinearity test, the three factors are significantly correlated with the absence of valid observations: (R2 = 0.481) and Std.Coef. (Std. Err.) are 0.306 (0.094), −0.453 (0.003) and −0.547 (0.019), respectively. Overall, these results highlight the substantial spatial heterogeneity of valid observations in China, confirming the reliability of the integration of Landsat-7 and 8 and Sentinel-2 imagery for paddy rice mapping based on phenology-based approaches. This can pave the way for a national-scale effort of rice mapping in China while further indicating potential omission errors in certain cloud-prone regions without sufficient optical observation data, i.e., the Sichuan Basin.

Inter-Annual Climate Variability Impact on Oil Palm Mapping

The contribution of oil palm plantations to the economic growth of tropical developing countries makes it essential to monitor their expansion into the tropical forest; consequently, most studies focus on improving mapping accuracy while using satellite imagery. However, accuracy can be hampered by atmospheric phenomena that can drastically change climatic conditions in tropical regions, affecting the spectral properties of the vegetation. In this sense, we studied the accuracy of palm plantation mapping by using features from different regions of the electromagnetic spectrum and a data fusion approach, and then compared the changes in accuracy over the years 2016, 2017, and 2018 (two of them with reported climatic anomalies). Optical-based maps obtained higher accuracy than thermal- and microwave-based maps, but they were the most affected by inter-annual climate variability (error margin between 5 and 10%), while thermal-based maps were the least affected (error margin between 8 and 9%). Data fusion combinations improved accuracy and reduced dissimilarities between years (e.g., phenology-based map accuracy changed by up to 20.8%, while phenology fused with microwave features changed by up to 6.8%). We conclude that inter-annual climate variability on land-cover mapping should be considered, especially if the outputs will be used as input in future studies.

Comparison of Methods for Reconstructing MODIS Land Surface Temperature under Cloudy Conditions

Land surface temperature (LST) is a vital parameter associated with the land–atmosphere interface. The Moderate Resolution Imaging Spectroradiometer (MODIS) LST product can provide precise LST with high time resolution, and is widely applied in various remote sensing temperature research. However, due to its inability to penetrate the cloud and fog, its quality is not able to meet the requirements of actual research. Hence, obtaining continuous and cloudless MODIS LST datasets remains challenging for researchers. The critical point is to reconstruct missing pixels. To compare the performance of different methods, first, three kinds of methods were used to reconstruct the missing pixels, namely, temporal, spatial, and spatiotemporal methods. The predicted values using these methods were validated by the automatic weather system data (AWS) in the Heihe river basin of China. The results demonstrated that, compared with other methods, linear temporal interpolation using Aqua data had the best performance in MODIS LST reconstruction in the Heihe river basin, with an RMSE of 7.13 K and an R2 of 0.82, and the NSE and PBias were 0.78 and −0.76%, respectively. Furthermore, the interpolation method was improved using adaptive windows and robust regression. First, the international Geosphere–Biosphere Program (IGBP) classification was employed to distinguish the different land surface types. Then, the invalid LST values were reconstructed using adjacent days’ effective LST values combined with a robust regression. Finally, a mean filter was applied to eliminate outliers. The overall results combined with ERA5 data were validated by AWS, with an RMSE of 6.96 K and an R2 of 0.79 and the NSE and PBias were 0.77 and −0.20%, respectively. The validation demonstrated that the scheme proposed in this paper is able to accurately reconstruct the missing values and improve the accuracy of the interpolation method to a certain extent when reconstructing MODIS LST.

Evaluating the Effectiveness of Machine Learning and Deep Learning Models Combined Time-Series Satellite Data for Multiple Crop Types Classification over a Large-Scale Region

Accurate extraction of crop cultivated area and spatial distribution is essential for food security. Crop classification methods based on machine learning and deep learning and remotely sensed time-series data are widely utilized to detect crop planting area. However, few studies assess the effectiveness of machine learning and deep learning algorithm integrated time-series satellite data for identifying multiple crop type classification over a large-scale region. Hence, this study aims to evaluate the effectiveness of machine learning and deep learning models in crop classification and provide a framework for large-scale multiple crop type classification based on time-series of satellite data. The time-series of the Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), and NaE (combined NDVI and EVI) were adopted as input features, and four widely used machine learning models, including Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbor (KNN), and their integration (Stacking) were compared to examine the performance of multiple crop types (i.e., wheat, corn, wheat-corn, early rice, and early rice-late rice) classification in the North China Plain. The performance of two types of deep learning classifiers: the One-dimensional Convolutional Neural Network (Conv1D) and Long Short-Term Memory Networks (LSTM), were also tested. The results showed that the NaE feature performed best among three input features, and the Stacking model produced the highest accuracy (77.12%) compared to other algorithms.

High-Resolution Mapping of Paddy Rice Extent and Growth Stages across Peninsular Malaysia Using a Fusion of Sentinel-1 and 2 Time Series Data in Google Earth Engine

Rice is the staple crop for more than half the world’s population, but there is a lack of high-resolution maps outlining rice areas and their growth stages. Most remote sensing studies map the rice extent; however, in tropical regions, rice is grown throughout the year with variable planting dates and cropping frequency. Thus, mapping rice growth stages is more useful than mapping only the extent. This study addressed this challenge by developing a phenology-based method. The hypothesis was that the unsupervised classification (k-means clustering) of Sentinel-1 and 2 time-series data could identify rice fields and growth stages, because (1) the presence of flooding during transplanting can be identified by Sentinel-1 VH backscatter; and (2) changes in the canopy of rice fields during growth stages (vegetative, generative, and ripening phases) up to the point of harvesting can be identified by Normalized Difference Vegetation Index (NDVI) time series. Using the proposed method, this study mapped rice field extent and cropping calendars across Peninsular Malaysia (131,598 km2) on the Google Earth Engine (GEE) platform. The Sentinel-1 and 2 monthly time series data from January 2019 to December 2020 were classified using k-means clustering to identify areas with similar phenological patterns. This approach resulted in 10-meter resolution maps of rice field extent, intensity, and cropping calendars. Validation using very high-resolution street view images from Google Earth showed that the predicted map had an overall accuracy of 95.95%, with a kappa coefficient of 0.92. In addition, the predicted crop calendars agreed well with the local government’s granary data. The results show that the proposed phenology-based method is cost-effective and can accurately map rice fields and growth stages over large areas. The information will be helpful in measuring the achievement of self-sufficiency in rice production and estimates of methane emissions from rice cultivation.

A Novel Spectral Index for Automatic Canola Mapping by Using Sentinel-2 Imagery

Because canola is a major oilseed crop, accurately determining its planting areas is crucial for ensuring food security and achieving UN 2030 sustainable development goals. However, when canola is extracted using remote-sensing data, winter wheat causes serious interference because it has a similar growth cycle and spectral reflectance characteristics. This interference seriously limits the classification accuracy of canola, especially in mixed planting areas. Here, a novel canola flower index (CFI) is proposed based on the red, green, blue, and near-infrared bands of Sentinel-2 images to improve the accuracy of canola mapping, based on the finding that spectral reflectance of canola on the red and green bands is higher than that of winter wheat during the canola flowering period. To investigate the potential of the CFI for extracting canola, the IsoData, support vector machine (SVM), and random forest (RF) classification methods were used to extract canola based on Sentinel-2 raw images and CFI images. The results show that the average overall accuracy and kappa coefficient based on CFI images were 94.77% and 0.89, respectively, which were 1.05% and 0.02, respectively, higher than those of the Sentinel-2 raw images. Then we found that a threshold of 0.14 on the CFI image could accurately distinguish canola from non-canola vegetation, which provides a solution for automatic mapping of canola. The overall classification accuracy and kappa coefficient of this threshold method were 96.02% and 0.92, which were very similar to those of the SVM and RF methods. Moreover, the advantage of the threshold classification method is that it reduces the dependence on training samples and has good robustness and high classification efficiency. Overall, this study shows that CFI and Sentinel-2 images provide a solution for automatic and accurate canola extraction.

Mapping Paddy Rice Distribution and Cropping Intensity in China from 2014 to 2019 with Landsat Images, Effective Flood Signals, and Google Earth Engine

Paddy rice cropping systems play a vital role in food security, water use, gas emission estimates, and grain yield prediction. Due to alterations in the labor structure and the high cost of paddy rice planting, the paddy rice cropping systems (single or double paddy rice) have drastically changed in China in recent years; many double-cropping paddy rice fields have been converted to single-cropping paddy rice or other crops, especially in southern China. Few maps detect single and double paddy rice and cropping intensity for paddy rice (CIPR) in China with a 30 m resolution. The Landsat-based and effective flooding signal-based phenology (EFSP) method, which distinguishes CIPR with the frequency of the effective flooding signal (EFe), was proposed and tested in China. The cloud/ice/shadow was excluded by bit arithmetic, generating a good observation map, and several non-paddy rice masks were established to improve the classification accuracy. Threshold values for single and double paddy rice were calculated through the mapped data and agricultural census data. Image processing (more than 684,000 scenes) and algorithm implementation were accomplished by a cloud computing approach with the Google Earth Engine (GEE) platform. The resultant maps of paddy rice from 2014 to 2019 were evaluated with data from statistical yearbooks and high-resolution images, with producer (user) accuracy and kappa coefficients ranging from 0.92 to 0.96 (0.76–0.87) and 0.67–0.80, respectively. Additionally, the determination coefficients for mapped and statistical data were higher than 0.88 from 2014 to 2019. Maps derived from EFSP illustrate that the single and double paddy rice systems are mainly concentrated in the Cfa (warm, fully humid, and hot summer, 49% vs. 56%) climate zone in China and show a slightly decreasing trend. The trend of double paddy rice is more pronounced than that of single paddy rice due to the high cost and shortages of rural household labor. However, single paddy rice fields expanded in Dwa (cold, dry winter, and hot summer, 11%) and Dwb (cold, dry winter, and warm summer, 9%) climate zones. The regional cropping intensity for paddy rice coincides with the paddy rice planting area but shows a significant decrease in south China, especially in Hunan Province, from 2014 to 2019. The results demonstrate that EFSP can effectively support the mapping of single and double paddy rice fields and CIPR in China, and the combinations of Landsat 7 and 8 provide enough good observations for EFSP to monitor paddy rice agriculture.

Large-Scale Rice Mapping Using Multi-Task Spatiotemporal Deep Learning and Sentinel-1 SAR Time Series

Timely and accurate cropland information at large spatial scales can improve crop management and support the government in decision making. Mapping the spatial extent and distribution of crops on a large spatial scale is challenging work due to the spatial variability. A multi-task spatiotemporal deep learning model, named LSTM-MTL, was developed in this study for large-scale rice mapping by utilizing time-series Sentinel-1 SAR data. The model showed a reasonable rice classification accuracy in the major rice production areas of the U.S. (OA = 98.3%, F1 score = 0.804), even when it only utilized SAR data. The model learned region-specific and common features simultaneously, and yielded a significant improved performance compared with RF and AtBiLSTM in both global and local training scenarios. We found that the LSTM-MTL model achieved a regional F1 score up to 10% higher than both global and local baseline models. The results demonstrated that the consideration of spatial variability via LSTM-MTL approach yielded an improved crop classification performance at a large spatial scale. We analyzed the input-output relationship through gradient backpropagation and found that low VH value in the early period and high VH value in the latter period were critical for rice classification. The results of in-season analysis showed that the model was able to yield a high accuracy (F1 score = 0.746) two months before rice maturity. The integration between multi-task learning and multi-temporal deep learning approach provides a promising approach for crop mapping at large spatial scales.

Mapping a Paddy Rice Area in a Cloudy and Rainy Region Using Spatiotemporal Data Fusion and a Phenology-Based Algorithm

The timely and accurate mapping of paddy rice is important to ensure food security and to protect the environment for sustainable development. Existing paddy rice mapping methods are often remote sensing technologies based on optical images. However, the availability of high-quality remotely sensed paddy rice growing area data is limited due to frequent cloud cover and rain over the southwest China. In order to overcome these limitations, we propose a paddy rice field mapping method by combining a spatiotemporal fusion algorithm and a phenology-based algorithm. First, a modified neighborhood similar pixel interpolator (MNSPI) time series approach was used to remove clouds on Sentinel-2 and Landsat 8 OLI images in 2020. A flexible spatiotemporal data fusion (FSDAF) model was used to fuse Sentinel-2 data and MODIS data to obtain multi-temporal Sentinel-2 images. Then, the fused remote sensing data were used to construct fusion time series data to produce time series vegetation indices (NDVI\LSWI) having a high spatiotemporal resolution (10 m and ≤16 days). On this basis, the unique physical characteristics of paddy rice during the transplanting period and other auxiliary data were combined to map paddy rice in Yongchuan District, Chongqing, China. Our results were validated by field survey data and showed a high accuracy of the proposed method indicated by an overall accuracy of 93% and the Kappa coefficient of 0.85. The paddy rice planting area map was also consistent with the official data of the third national land survey; at the town level, the correlation between official survey data and paddy rice area was 92.5%. The results show that this method can effectively map paddy rice fields in a cloudy and rainy area.

Evaluating the performance of eight drought indices for capturing soil moisture dynamics in various vegetation regions over China

Drought is pervasive global hazard and seriously impacts ecology. Particularly, vegetation drought, which is chiefly driven by soil moisture (SM) deficiency, has a direct bearing on grain production and human livelihoods. Various drought indices associated with vegetation and SM conditions have been proposed to monitor and detect vegetation drought. In this study, we evaluated the performance of eight drought indices, including Drought Severity Index (DSI), Evaporation Stress Index (ESI), Normalized Vegetation Supply Water Index (NVSWI), Temperature-Vegetation Dryness Index (TVDI), Temperature Vegetation Precipitation Dryness Index (TVPDI), Vegetation Health Index (VHI), Self-calibrating Palmer Drought Severity Index (SC-PDSI) and Standardized Precipitation Evapotranspiration Index (SPEI), for capturing SM dynamic (derived from Copernicus Climate Change Service) across the six main vegetation coverage types of China. Our results showed DSI and ESI had the best overall performance. When exploring the reasons for the uncertainty of these indices (except SC-PDSI and SPEI) in the evaluation, we found that, in the non-arable regions, the time lag effect of drought indices on SM, the average state and rangeability of corresponding variables and the climatic conditions (precipitation and temperature) all impacted the performance of DSI, ESI, NVSWI, TVPDI and VHI. In the arable region, cropland types (paddy field and non-paddy field) and the uncertainty of SM data mainly caused the uncertainties of the above five indices. With regard to the TVDI, abnormalities of dry and wet edges fitting may be the primary factor affecting its performance. These results demonstrated that these drought indices with reliable and robust performance of capturing SM dynamics can be suggested to characterize the trend of SM. Certainly, this study can provide a reference for the improvement of existing drought indices and the establishment of new drought indices.

A Review of Reconstructing Remotely Sensed Land Surface Temperature under Cloudy Conditions

Land surface temperature (LST) is an important environmental parameter in climate change, urban heat islands, drought, public health, and other fields. Thermal infrared (TIR) remote sensing is the main method used to obtain LST information over large spatial scales. However, cloud cover results in many data gaps in remotely sensed LST datasets, greatly limiting their practical applications. Many studies have sought to fill these data gaps and reconstruct cloud-free LST datasets over the last few decades. This paper reviews the progress of LST reconstruction research. A bibliometric analysis is conducted to provide a brief overview of the papers published in this field. The existing reconstruction algorithms can be grouped into five categories: spatial gap-filling methods, temporal gap-filling methods, spatiotemporal gap-filling methods, multi-source fusion-based gap-filling methods, and surface energy balance-based gap-filling methods. The principles, advantages, and limitations of these methods are described and discussed. The applications of these methods are also outlined. In addition, the validation of filled LST values’ cloudy pixels is an important concern in LST reconstruction. The different validation methods applied for reconstructed LST datasets are also reviewed herein. Finally, prospects for future developments in LST reconstruction are provided.

Deep Learning-Based Phenological Event Modeling for Classification of Crops

Classification of crops using time-series vegetation index (VI) curves requires appropriate modeling of phenological events and their characteristics. The current study explores the use of capsules, a group of neurons having an activation vector, to learn the characteristic features of the phenological curves. In addition, joint optimization of denoising and classification is adopted to improve the generalizability of the approach and to make it resilient to noise. The proposed approach employs reconstruction loss as a regularizer for classification, whereas the crop-type label is used as prior information for denoising. The activity vector of the class capsule is applied to sample the latent space conditioned on the cell state of a Long Short-Term Memory (LSTM) that integrates the sequences of the phenological events. Learning of significant phenological characteristics is facilitated by adversarial variational encoding in conjunction with constraints to regulate latent representations and embed label information. The proposed architecture, called the variational capsule network (VCapsNet), significantly improves the classification and denoising results. The performance of VCapsNet can be attributed to the suitable modeling of phenological events and the resilience to outliers and noise. The maxpooling-based capsule implementation yields better results, particularly with limited training samples, compared to the conventional implementations. In addition to the confusion matrix-based accuracy measures, this study illustrates the use of interpretability-based evaluation measures. Moreover, the proposed approach is less sensitive to noise and yields good results, even at shallower depths, compared to the main existing approaches. The performance of VCapsNet in accurately classifying wheat and barley crops indicates that the approach addresses the issues in crop-type classification. The approach is generic and effectively models the crop-specific phenological features and events. The interpretability-based evaluation measures further indicate that the approach successfully identifies the crop transitions, in addition to the planting, heading, and harvesting dates. Due to its effectiveness in crop-type classification, the proposed approach is applicable to acreage estimation and other applications in different scales.

Mapping Croplands in the Granary of the Tibetan Plateau Using All Available Landsat Imagery, A Phenology-Based Approach, and Google Earth Engine

The Tibetan Plateau (TP), known as “The Roof of World”, has expansive alpine grasslands and is a hotspot for climate change studies. However, cropland expansion and increasing anthropogenic activities have been poorly documented, let alone the effects of agricultural activities on food security and environmental change in the TP. The existing cropland mapping products do not depict the spatiotemporal characteristics of the TP due to low accuracies and inconsistent cropland distribution, which is affected by complicated topography and impedes our understanding of cropland expansion and its associated environmental impacts. One of the biggest challenges of cropland mapping in the TP is the diverse crop phenology across a wide range of elevations. To decrease the classification errors due to elevational differences in crop phenology, we developed two pixel- and phenology-based algorithms to map croplands using Landsat imagery and the Google Earth Engine platform along the Brahmaputra River and its two tributaries (BRTT) in the Tibet Autonomous Region, also known as the granary of TP, in 2015–2019. Our first phenology-based cropland mapping algorithm (PCM1) used different thresholds of land surface water index (LSWI) by considering varied crop phenology along different elevations. The second algorithm (PCM2) further offsets the phenological discrepancy along elevational gradients by considering the length and peak of the growing season. We found that PCM2 had a higher accuracy with fewer images compared with PCM1. The number of images for PCM2 was 279 less than PCM1, and the Matthews correlation coefficient for PCM2 was 0.036 higher than PCM1. We also found that the cropland area in BRTT was estimated to be 1979 ± 52 km2 in the late 2010s. Croplands were mainly distributed in the BRTT basins with elevations of 3800–4000 m asl. Our phenology-based methods were effective for mapping croplands in mountainous areas. The spatially explicit information on cropland area and distribution in the TP aid future research into the effects of cropland expansion on food security and environmental change in the TP.

Testing the Contribution of Multi-Source Remote Sensing Features for Random Forest Classification of the Greater Amanzule Tropical Peatland

Tropical peatlands such as Ghana's Greater Amanzule peatland are highly valuable ecosystems and under great pressure from anthropogenic land use activities. Accurate measurement of their occurrence and extent is required to facilitate sustainable management. A key challenge, however, is the high cloud cover in the tropics that limits optical remote sensing data acquisition. In this work we combine optical imagery with radar and elevation data to optimise land cover classification for the Greater Amanzule tropical peatland. Sentinel-2, Sentinel-1 and Shuttle Radar Topography Mission (SRTM) imagery were acquired and integrated to drive a machine learning land cover classification using a random forest classifier. Recursive feature elimination was used to optimize high-dimensional and correlated feature space and determine the optimal features for the classification. Six datasets were compared, comprising different combinations of optical, radar and elevation features. Results showed that the best overall accuracy (OA) was found for the integrated Sentinel-2, Sentinel-1 and SRTM dataset (S2+S1+DEM), significantly outperforming all the other classifications with an OA of 94%. Assessment of the sensitivity of land cover classes to image features indicated that elevation and the original Sentinel-1 bands contributed the most to separating tropical peatlands from other land cover types. The integration of more features and the removal of redundant features systematically increased classification accuracy. We estimate Ghana's Greater Amanzule peatland covers 60,187 ha. Our proposed methodological framework contributes a robust workflow for accurate and detailed landscape-scale monitoring of tropical peatlands, while our findings provide timely information critical for the sustainable management of the Greater Amanzule peatland.

A Novel Index to Detect Vegetation in Urban Areas Using UAV-Based Multispectral Images

Unmanned aerial vehicles (UAVs) equipped with high-resolution multispectral cameras have increasingly been used in urban planning, landscape management, and environmental monitoring as an important complement to traditional satellite remote sensing systems. Interest in urban regeneration projects is on the rise in Korea, and the results of UAV-based urban vegetation analysis are in the spotlight as important data to effectively promote urban regeneration projects. Vegetation indices have been used to obtain vegetation information in a wide area using the multispectral bands of satellites. UAV images have recently been used to obtain vegetation information in a more rapid and precise manner. In this study, multispectral images were acquired using a UAV equipped with a Micasense RedEde MX camera to analyze vegetation indices, such as the Normalized Difference Vegetation Index (NDVI), Green Normalized Difference Vegetation Index (GNDVI), Blue Normalized Difference Vegetation Index (BNDVI), Red Green Blue Vegetation Index (RGBVI), Green Red Vegetation Index (GRVI), and Soil Adjusted Vegetation Index (SAVI). However, in the process of analyzing urban vegetation using the existing vegetation indices, it became clear that the vegetation index values of long-run steel roofing, waterproof coated roofs, and urethane-coated areas are often similar to, or slightly higher than, those of grass. In order to improve the problem of misclassification of vegetation, various equations were tested by combining multispectral bands. Kappa coefficient analysis showed that the squared Red-Blue NDVI index produced the best results when analyzing vegetation reflecting urban land cover. The novel vegetation index developed in this study will be very useful for effective analysis of vegetation in urban areas with various types of land cover, such as long-run steel roofing, waterproof coated roofs, and urethane-coated areas.

Mapping abandoned farmland in China using time series MODIS NDVI

Farmland abandonment is an important aspect of land cover land use (LCLU) that has both positive and negative effects on the environment. There is limited information regarding the spatial distribution of abandoned farmland in China. In this study, we 1) use a C5.0 decision tree classification based on a 250 m spatial resolution Moderate-resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Index (NDVI) time series to develop LCLU maps for the period 2000-2017; 2) produce time series maps of the abandoned farmland during the period 2002-2017 based on LCLU maps; and 3) analyze spatial distribution, intensity, trend, frequency, and recultivation in terms of farmland abandonment. The results show that the overall accuracy of the LCLU maps ranged from 85.45% to 90.56% over the study period (2002-2017). The average national rate of farmland abandonment was 5% over the period 2002-2017, with the lowest rate in 2002 and the highest in 2015. Approximately 60% of the 34 agricultural areas in China showed an increasing trend for abandonment including some of the main grain-producing areas such as the North China Plain and the middle and lower reaches of the Yangtze River. Abandoned farmland areas with an abandonment frequency of ≤ 3 y comprised 58.24% of the total area of abandoned farmland. The first map detailing the distribution of the abandoned farmland across China was produced in this study, providing guidance for the development of a method of identifying abandoned farmland on a large spatial scale.

Rice Mapping and Growth Monitoring Based on Time Series GF-6 Images and Red-Edge Bands

Accurate rice mapping and growth monitoring are of great significance for ensuring food security and agricultural sustainable development. Remote sensing (RS), as an efficient observation technology, is expected to be useful for rice mapping and growth monitoring. Due to the fragmented distribution of paddy fields and the undulating terrain in Southern China, it is very difficult in rice mapping. Moreover, there are many crops with the same growth period as rice, resulting in low accuracy of rice mapping. We proposed a red-edge decision tree (REDT) method based on the combination of time series GF-6 images and red-edge bands to solve this problem. The red-edge integral and red-edge vegetation index integral were computed by using two red-edge bands derived from GF-6 images to construct the REDT. Meanwhile, the conventional method based on time series normalized difference vegetation index (NDVI), normalized difference water index (NDWI), enhanced vegetation index (EVI) (NNE) was employed to compare the effectiveness of rice mapping. The results indicated that the overall accuracy and Kappa coefficient of REDT ranged from 91%–94% and 0.82–0.87, improving about 7% and 0.15 compared with the NNE method. This proved that the proposed technology was able to efficiently solve the problem of rice mapping on a large scale and regions with fragmented landscapes. Additionally, two red-edge bands of GF-6 images were applied to monitor rice growth. It concluded that the two red-edge bands played different roles in rice growth monitoring. The red-edge bands of GF-6 images were superior in rice mapping and growth monitoring. Further study needs to develop more vegetation indices (VIs) related to the red-edge to make the best use of red-edge characteristics in precision agriculture.

Mapping Paddy Rice with Satellite Remote Sensing: A Review

Paddy rice is a staple food of three billion people in the world. Timely and accurate estimation of the paddy rice planting area and paddy rice yield can provide valuable information for the government, planners and decision makers to formulate policies. This article reviews the existing paddy rice mapping methods presented in the literature since 2010, classifies these methods, and analyzes and summarizes the basic principles, advantages and disadvantages of these methods. According to the data sources used, the methods are divided into three categories: (I) Optical mapping methods based on remote sensing; (II) Mapping methods based on microwave remote sensing; and (III) Mapping methods based on the integration of optical and microwave remote sensing. We found that the optical remote sensing data sources are mainly MODIS, Landsat, and Sentinel-2, and the emergence of Sentinel-1 data has promoted research on radar mapping methods for paddy rice. Multisource data integration further enhances the accuracy of paddy rice mapping. The best methods are phenology algorithms, paddy rice mapping combined with machine learning, and multisource data integration. Innovative methods include the time series similarity method, threshold method combined with mathematical models, and object-oriented image classification. With the development of computer technology and the establishment of cloud computing platforms, opportunities are provided for obtaining large-scale high-resolution rice maps. Multisource data integration, paddy rice mapping under different planting systems and the connection with global changes are the focus of future development priorities.

An Approach to High-Resolution Rice Paddy Mapping Using Time-Series Sentinel-1 SAR Data in the Mun River Basin, Thailand

Timely and accurate regional rice paddy monitoring plays a significant role in maintaining the sustainable rice production, food security, and agricultural development. This study proposes an operational automatic approach to mapping rice paddies using time-series SAR data. The proposed method integrates time-series Sentinel-1 data, auxiliary data of global surface water, and rice phenological characteristics with Google Earth Engine cloud computing platform. A total of 402 Sentinel-1 scenes from 2017 were used for mapping rice paddies extent in the Mun River basin. First, the calculated minimum and maximum values of the backscattering coefficient of permanent water (a classification type within global surface water data) in a year was used as the threshold range for extracting the potential extent. Then, three rice phenological characteristics were extracted based on the time-series curve of each pixel, namely the date of the beginning of the season (DBS), date of maximum backscatter during the peak growing season (DMP), and length of the vegetative stage (LVS). After setting a threshold for each phenological parameter, the final rice paddy extent was identified. Rice paddy map produced in this study was highly accurate and agreed well with field plot data and rice map products from the International Rice Research Institute (IRRI). The results had a total accuracy of 89.52% and an F1 score of 0.91, showing that the spatiotemporal pattern of extracted rice cover was consistent with ground truth samples in the Mun River basin. This approach could be expanded to other rice-growing regions at the national scale, or even the entire Indochina Peninsula and Southeast Asia.

Combination of Landsat 8 OLI and Sentinel-1 SAR Time-Series Data for Mapping Paddy Fields in Parts of West and Central Java Provinces, Indonesia

The rise of Google Earth Engine, a cloud computing platform for spatial data, has unlocked seamless integration for multi-sensor and multi-temporal analysis, which is useful for the identification of land-cover classes based on their temporal characteristics. Our study aims to employ temporal patterns from monthly-median Sentinel-1 (S1) C-band synthetic aperture radar data and cloud-filled monthly spectral indices, i.e., Normalized Difference Vegetation Index (NDVI), Modified Normalized Difference Water Index (MNDWI), and Normalized Difference Built-up Index (NDBI), from Landsat 8 (L8) OLI for mapping rice cropland areas in the northern part of Central Java Province, Indonesia. The harmonic function was used to fill the cloud and cloud-masked values in the spectral indices from Landsat 8 data, and smile Random Forests (RF) and Classification And Regression Trees (CART) algorithms were used to map rice cropland areas using a combination of monthly S1 and monthly harmonic L8 spectral indices. An additional terrain variable, Terrain Roughness Index (TRI) from the SRTM dataset, was also included in the analysis. Our results demonstrated that RF models with 50 (RF50) and 80 (RF80) trees yielded better accuracy for mapping the extent of paddy fields, with user accuracies of 85.65% (RF50) and 85.75% (RF80), and producer accuracies of 91.63% (RF80) and 93.48% (RF50) (overall accuracies of 92.10% (RF80) and 92.47% (RF50)), respectively, while CART yielded a user accuracy of only 84.83% and a producer accuracy of 80.86%. The model variable importance in both RF50 and RF80 models showed that vertical transmit and horizontal receive (VH) polarization and harmonic-fitted NDVI were identified as the top five important variables, and the variables representing February, April, June, and December contributed more to the RF model. The detection of VH and NDVI as the top variables which contributed up to 51% of the Random Forest model indicated the importance of the multi-sensor combination for the identification of paddy fields.

Automatic Mapping of Rice Growth Stages Using the Integration of SENTINEL-2, MOD13Q1, and SENTINEL-1

Rice (Oryza sativa L.) is a staple food crop for more than half of the world’s population. Rice production is facing a myriad of problems, including water shortage, climate, and land-use change. Accurate maps of rice growth stages are critical for monitoring rice production and assessing its impacts on national and global food security. Rice growth stages are typically monitored by coarse-resolution satellite imagery. However, it is difficult to accurately map due to the occurrence of mixed pixels in fragmented and patchy rice fields, as well as cloud cover, particularly in tropical countries. To solve these problems, we developed an automated mapping workflow to produce near real-time multi-temporal maps of rice growth stages at a 10-m spatial resolution using multisource remote sensing data (Sentinel-2, MOD13Q1, and Sentinel-1). This study was investigated between 1 June and 29 September 2018 in two (wet and dry) areas of Java Island in Indonesia. First, we built prediction models based on Sentinel-2, and fusion of MOD13Q1/Sentinel-1 using the ground truth information. Second, we applied the prediction models on all images in area and time and separation between the non-rice planting class and rice planting class over the cropping pattern. Moreover, the model’s consistency on the multitemporal map with a 5–30-day lag was investigated. The result indicates that the Sentinel-2 based model classification gives a high overall accuracy of 90.6% and the fusion model MOD13Q1/Sentinel-1 shows 78.3%. The performance of multitemporal maps was consistent between time lags with an accuracy of 83.27–90.39% for Sentinel-2 and 84.15% for the integration of Sentinel-2/MOD13Q1/Sentinel-1. The results from this study show that it is possible to integrate multisource remote sensing for regular monitoring of rice phenology, thereby generating spatial information to support local-, national-, and regional-scale food security applications.

Quantifying expansion and removal of Spartina alterniflora on Chongming island, China, using time series Landsat images during 1995-2018

The rampant encroachment of Spartina alterniflora into coastal wetlands of China over the past decades has adversely affected both coastal ecosystems and socio-economic systems. However, there are no annual or multi-year epoch maps of Spartina saltmarsh in China, which hinders our understanding and management of Spartina invasion. In this study, we selected Chongming island, China, where Spartina saltmarsh had expanded rapidly since its introduction in the 1990s. We investigated phenology of Spartina, Phragmites and Scirpus saltmarshes, and the time series vegetation indices derived from Landsat images showed that Spartina saltmarsh did not green-up in April-May and stayed green in December-January, which differed from the phenology of Phragmites and Scirpus saltmarshes. We developed a pixel- and phenology-based algorithm that used time series Landsat data to identify and map Spartina saltmarsh, and we applied it to quantify the temporal dynamics (expansion and removal) of Spartina saltmarsh on Chongming island during 1995-2018. The resultant maps showed that Spartina saltmarsh area on Chongming island increased from ~4 ha in 1995 to ~2,067 ha in 2012 but dropped substantially to ~729 ha in 2016 after a large-scale ecological engineering project (US$ 186 million) was started to remove Spartina during 2013-2016. Chongming island still had ~1,315 ha Spartina saltmarsh in 2018, and majority of it was distributed outside the Chongming Dongtan National Nature Reserve, which could serve as the sources for reinvasion in the near future. This study demonstrates the feasibility of using time series Landsat images, pixel- and phenology-based algorithm, and GEE platform to identify and map Spartina saltmarsh over years in the region, which is useful to the management of invasive plants in coastal wetlands.

A Soft Computing Approach for Selecting and Combining Spectral Bands

We introduce a soft computing approach for automatically selecting and combining indices from remote sensing multispectral images that can be used for classification tasks. The proposed approach is based on a Genetic-Programming (GP) framework, a technique successfully used in a wide variety of optimization problems. Through GP, it is possible to learn indices that maximize the separability of samples from two different classes. Once the indices specialized for all the pairs of classes are obtained, they are used in pixelwise classification tasks. We used the GP-based solution to evaluate complex classification problems, such as those that are related to the discrimination of vegetation types within and between tropical biomes. Using time series defined in terms of the learned spectral indices, we show that the GP framework leads to superior results than other indices that are used to discriminate and classify tropical biomes.

Mapping Rice Paddy Based on Machine Learning with Sentinel-2 Multi-Temporal Data: Model Comparison and Transferability

Rice is an important agricultural crop in the Southwest Hilly Area, China, but there has been a lack of efficient and accurate monitoring methods in the region. Recently, convolutional neural networks (CNNs) have obtained considerable achievements in the remote sensing community. However, it has not been widely used in mapping a rice paddy, and most studies lack the comparison of classification effectiveness and efficiency between CNNs and other classic machine learning models and their transferability. This study aims to develop various machine learning classification models with remote sensing data for comparing the local accuracy of classifiers and evaluating the transferability of pretrained classifiers. Therefore, two types of experiments were designed: local classification experiments and model transferability experiments. These experiments were conducted using cloud-free Sentinel-2 multi-temporal data in Banan District and Zhongxian County, typical hilly areas of Southwestern China. A pure pixel extraction algorithm was designed based on land-use vector data and a Google Earth Online image. Four convolutional neural network (CNN) algorithms (one-dimensional (Conv-1D), two-dimensional (Conv-2D) and three-dimensional (Conv-3D_1 and Conv-3D_2) convolutional neural networks) were developed and compared with four widely used classifiers (random forest (RF), extreme gradient boosting (XGBoost), support vector machine (SVM) and multilayer perceptron (MLP)). Recall, precision, overall accuracy (OA) and F1 score were applied to evaluate classification accuracy. The results showed that Conv-2D performed best in local classification experiments with OA of 93.14% and F1 score of 0.8552 in Banan District, OA of 92.53% and F1 score of 0.8399 in Zhongxian County. CNN-based models except Conv-1D provided more desirable performance than non-CNN classifiers. Besides, among the non-CNN classifiers, XGBoost received the best result with OA of 89.73% and F1 score of 0.7742 in Banan District, SVM received the best result with OA of 88.57% and F1 score of 0.7538 in Zhongxian County. In model transferability experiments, almost all CNN classifiers had low transferability. RF and XGBoost models have achieved acceptable F1 scores for transfer (RF = 0.6673 and 0.6469, XGBoost = 0.7171 and 0.6709, respectively).

Satellite-Based Observations Reveal Effects of Weather Variation on Rice Phenology

Obtaining detailed data on the spatio-temporal variation in crop phenology is critical to increasing our understanding of agro-ecosystem function, such as their response to weather variation and climate change. It is challenging to collect such data over large areas through field observations. The use of satellite remote sensing data has made phenology data collection easier, although the quality and the utility of such data to understand agro-ecosystem function have not been widely studied. Here, we evaluated satellite data-based estimates of rice phenological stages in California, USA by comparing them with survey data and with predictions by a temperature-driven phenology model. We then used the satellite data-based estimates to quantify the crop phenological response to changes in weather. We used time-series of MODIS satellite data and PhenoRice, a rule-based rice phenology detection algorithm, to determine annual planting, heading and harvest dates of paddy rice in California between 2002 and 2017. At the state level, our satellite-based estimates of rice phenology were very similar to the official survey data, particularly for planting and harvest dates (RMSE = 3.8–4.0 days). Satellite based observations were also similar to predictions by the DD10 temperature-driven phenology model. We analyzed how the timing of these phenological stages varied with concurrent temperature and precipitation over this 16-year time period. We found that planting was earlier in warm springs (−1.4 days °C−1 for mean temperature between mid-April and mid-May) and later in wet years (5.3 days 100 mm-1 for total precipitation from March to April). Higher mean temperature during the pre-heading period of the growing season advanced heading by 2.9 days °C−1 and shortened duration from planting to heading by 1.9 days °C−1. The entire growing season was reduced by 3.2 days °C−1 because of the increased temperature during the rice season. Our findings confirm that satellite data can be an effective way to estimate variations in rice phenology and can provide critical information that can be used to improve understanding of agricultural responses to weather variation.

Large increases of paddy rice area, gross primary production, and grain production in Northeast China during 2000-2017

China is the largest rice producer and consumer in the world. Accurate estimations of paddy rice planting area and rice grain production is important for feeding the increasing population in China. However, Southern China had substantial losses in paddy rice area over the last three decades in those regions where paddy rice has traditionally been produced. Several studies have shown increased paddy rice area in Northeast China. Here we document the annual dynamics of paddy rice area, gross primary production (GPP), and grain production in Northeast China (Heilongjiang, Jilin and Liaoning provinces) during 2000-2017 using agricultural statistical data, satellite images, and model simulations. Annual maps derived from satellite images show that paddy rice area in Northeast China has increased by 3.68 million ha from 2000 to 2017, which is more than the total combined paddy rice area of North Korea, South Korea, and Japan. Approximately 82% of paddy rice pixels had an increase in annual GPP during 2000-2017. The expansion of paddy rice area slowed down substantially since 2015. Annual GPP from those paddy rice fields cultivated continuously over the 18 years were moderately higher than that from other paddy rice fields, which suggested that improved management practices could increase grain production in the region. There was a strong linear relationship between annual GPP and annual rice grain production in Northeast China by province and year, which illustrates the potential of using satellite-based data-driven model to track and assess grain production of paddy rice in the region. Northeast China is clearly an emerging rice production base and plays an increasing role in crop production and food security in China. However, many challenges for the further expansion and sustainable cultivation of paddy rice in Northeast China remain.

Tracking annual changes of coastal tidal flats in China during 1986-2016 through analyses of Landsat images with Google Earth Engine

Tidal flats (non-vegetated area), along with coastal vegetation area, constitute the coastal wetlands (intertidal zone) between high and low water lines, and play an important role in wildlife, biodiversity and biogeochemical cycles. However, accurate annual maps of coastal tidal flats over the last few decades are unavailable and their spatio-temporal changes in China are unknown. In this study, we analyzed all the available Landsat TM/ETM+/OLI imagery (~ 44,528 images) using the Google Earth Engine (GEE) cloud computing platform and a robust decision tree algorithm to generate annual frequency maps of open surface water body and vegetation to produce annual maps of coastal tidal flats in eastern China from 1986 to 2016 at 30-m spatial resolution. The resulting map of coastal tidal flats in 2016 was evaluated using very high-resolution images available in Google Earth. The total area of coastal tidal flats in China in 2016 was about 731,170 ha, mostly distributed in the provinces around Yellow River Delta and Pearl River Delta. The interannual dynamics of coastal tidal flats area in China over the last three decades can be divided into three periods: a stable period during 1986-1992, an increasing period during 1993-2001 and a decreasing period during 2002-2016. The resulting annual coastal tidal flats maps could be used to support sustainable coastal zone management policies that preserve coastal ecosystem services and biodiversity in China.

Large-Scale Marsh Loss Reconstructed from Satellite Data in the Small Sanjiang Plain since 1965: Process, Pattern and Driving Force

Monitoring wetland dynamics and related land-use changes over long-time periods is essential to understanding wetland evolution and supporting knowledge-based conservation policies. Combining multi-source remote sensing images, this study identifies the dynamics of marshes, a core part of wetlands, in the Small Sanjiang Plain (SSP), from 1965 to 2015. The influence of human activities on marsh patterns is estimated quantitatively by the trajectory analysis method. The results indicate that the marsh area decreased drastically by 53.17% of the total SSP area during the study period, which covered the last five decades. The marsh mostly transformed to paddy field and dry farmland in the SSP from 1965 to 2015, indicating that agricultural encroachment was the dominant contributor to marsh degradation in the area. Analysis of the landscape indexes indicates that marsh fragmentation was aggravated during the past five decades in the SSP. Trajectory analysis also indicated that human activities have acted as the primary driving force of marsh changes in the SSP since 1965. This study provides scientific information to better understand the evolution of the wetland and to implement ecological conservation and sustainable management of the wetlands in the future.

Fingerprint of rice paddies in spatial-temporal dynamics of atmospheric methane concentration in monsoon Asia

Agriculture (e.g., rice paddies) has been considered one of the main emission sources responsible for the sudden rise of atmospheric methane concentration (XCH₄) since 2007, but remains debated. Here we use satellite-based rice paddy and XCH₄ data to investigate the spatial–temporal relationships between rice paddy area, rice plant growth, and XCH₄ in monsoon Asia, which accounts for ~87% of the global rice area. We find strong spatial consistencies between rice paddy area and XCH₄ and seasonal consistencies between rice plant growth and XCH₄. Our results also show a decreasing trend in rice paddy area in monsoon Asia since 2007, which suggests that the change in rice paddy area could not be one of the major drivers for the renewed XCH₄ growth, thus other sources and sinks should be further investigated. Our findings highlight the importance of satellite-based paddy rice datasets in understanding the spatial–temporal dynamics of XCH₄ in monsoon Asia.

The role of paddy rice agriculture in the spatial and temporal dynamics of atmospheric methane concentration remains unclear. Here, Zhang et al. show that regions with dense rice paddies have high satellite-based column averaged CH₄ concentrations (XCH₄), and that seasonal dynamics of XCH₄ mirror those of paddy rice growth.

Optimizing Feature Selection of Individual Crop Types for Improved Crop Mapping

Accurate crop planting area information is of significance for understanding regional food security and agricultural development planning. While increasing numbers of medium resolution satellite imagery and improved classification algorithms have been used for crop mapping, limited efforts have been made in feature selection, despite its vital impacts on crop classification. Furthermore, different crop types have their unique spectral and phenology characteristics; however, the different features of individual crop types have not been well understood and considered in previous studies of crop mapping. Here, we examined an optimized strategy to integrate specific features of individual crop types for mapping an improved crop type layer in the Sanjiang Plain, a new food bowl in China, by using all Sentinel-2 time series images in 2018. First, an automatic spectro-temporal feature selection (ASTFS) method was used to obtain optimal features for individual crops (rice, corn, and soybean), including sorting all features by the global separability indices for each crop and removing redundant features by accuracy changes when adding new features. Second, the ASTFS-based optimized feature sets for individual crops were used to produce three crop probability maps with the Random Forest classifier. Third, the probability maps were then composited into the final crop layer by considering the probability of each crop at every pixel. The resultant crop layer showed an improved accuracy (overall accuracy = 93.94%, Kappa coefficient = 0.92) than the other classifications without such a feature optimizing process. Our results indicate the potential of the ASTFS method for improving regional crop mapping.

Climate variability, satellite-derived physical environmental data and human leptospirosis: A retrospective ecological study in China

BACKGROUND

In the past three decades, the incidence rate of notified leptospirosis cases in China have steeply declined and are now circumscribed to discrete areas in the country. Previous research showed that climate and environmental variation may play an important role in leptospirosis transmission. However, quantitative associations between climate, environmental factors and leptospirosis in the high-risk areas in China, is still poorly understood.

OBJECTIVE

To quantify the temporal effects of climate and remotely-sensed physical environmental factors on human leptospirosis in the high-risk counties in China.

METHODS

Time series seasonal decomposition was performed to explore the seasonality pattern of leptospirosis incidence in Mengla County, Yunnan and Yilong County, Sichuan for the period 2006-2016. Time series cross-correlation analysis was carried out to examine lagged effects of rainfall, relative humidity, normalized difference vegetation index (NDVI), modified normalized difference water index (MNDWI) and land surface temperature (LST) on leptospirosis. The associations of climatic and physical environment factors with leptospirosis in each county were assessed by using a generalized linear regression model with negative binomial link, adjusted by seasonal components.

RESULTS

Leptospirosis incidence in both counties showed strong and unique annual seasonality. Our results show that in Mengla County leptospirosis notifications exhibits a bi-modal temporal pattern while in Yilong County it follows a typical single epidemic curve. After adjusting for seasonality, the final best-fitting model for Mengla County indicated that leptospirosis notifications were significantly associated with present LST values (incidence rate ratio, IRR = 0.857, 95% confidence interval (CI):0.729-0.929) and rainfall at a lag of 6-months (IRR = 0.989; 95% CI: 0.985-0.993). The incidence of leptospirosis in Yilong was associated with rainfall at 1-month lag (IRR = 1.013, 95% CI: 1.003-1.023), LST (3-months lag) (IRR = 1.193, 95% CI: 1.095-1.301), and MNDWI (5-months lag) (IRR = 7.960, 95% CI: 1.241-47.66).

CONCLUSIONS

Our study identified lagged effects between leptospirosis incidence and climate and remotely-sensed environmental factors in the two most endemic counties in China. Rainfall in combination with satellite derived physical environment factors provided better insight of the local epidemiology as well as good predictors for leptospirosis outbreak in both counties. This would also be an avenue for the development of leptospirosis early warning systems to support leptospirosis control in China.

Mapping Paddy Rice Planting Area in Northeastern China Using Spatiotemporal Data Fusion and Phenology-Based Method

Accurate paddy rice mapping with fine spatial detail is significant for ensuring food security and maintaining sustainable environmental development. In northeastern China, rice is planted in fragmented and patchy fields and its production has reached over 10% of the total amount of rice production in China, which has brought the increasing need for updated paddy rice maps in the region. Existing methods for mapping paddy rice are often based on remote sensing techniques by using optical images. However, it is difficult to obtain high quality time series remote sensing data due to the frequent cloud cover in rice planting area and low temporal sampling frequency of satellite imagery. Therefore, paddy rice maps are often developed using few Landsat or time series MODIS images, which has limited the accuracy of paddy rice mapping. To overcome these limitations, we presented a new strategy by integrating a spatiotemporal fusion algorithm and phenology-based algorithm to map paddy rice fields. First, we applied the spatial and temporal adaptive reflectance fusion model (STARFM) to fuse the Landsat and MODIS data and obtain multi-temporal Landsat-like images. From the fused Landsat-like images and the original Landsat images, we derived time series vegetation indices (VIs) with high temporal and high spatial resolution. Then, the phenology-based algorithm, considering the unique physical features of paddy rice during the flooding and transplanting phases/open-canopy period, was used to map paddy rice fields. In order to prove the effectiveness of the proposed strategy, we compared our results with those from other three classification strategies: (1) phenology-based classification based on original Landsat images only, (2) phenology-based classification based on original MODIS images only and (3) random forest (RF) classification based on both Landsat and Landsat-like images. The validation experiments indicate that our fusion-and phenology-based strategy could improve the overall accuracy of classification by 6.07% (from 92.12% to 98.19%) compared to using Landsat data only, and 8.96% (from 89.23% to 98.19%) compared to using MODIS data, and 4.66% (from93.53% to 98.19%) compared to using the RF algorithm. The results show that our new strategy, by integrating the spatiotemporal fusion algorithm and phenology-based algorithm, can provide an effective and robust approach to map paddy rice fields in regions with limited available images, as well as the areas with patchy and fragmented fields.

Automated Near-Real-Time Mapping and Monitoring of Rice Extent, Cropping Patterns, and Growth Stages in Southeast Asia Using Sentinel-1 Time Series on a Google Earth Engine Platform

More than 50% of the world’s population consumes rice. Accurate and up-to-date information on rice field extent is important to help manage food and water security. Currently, field surveys or MODIS satellite data are used to estimate rice growing areas. This study presents a cost-effective methodology for near-real-time mapping and monitoring of rice growth extent and cropping patterns over a large area. This novel method produces high-resolution monthly maps (10 m resolution) of rice growing areas, as well as rice growth stages. The method integrates temporal Sentinel-1 data and rice phenological parameters with the Google Earth Engine (GEE) cloud-based platform. It uses monthly median time series of Sentinel-1 at VH polarization from September 2016 to October 2018. The two study areas are the northern region of West Java, Indonesia (0.75 million ha), and the Kedah and Perlis states in Malaysia (over 1 million ha). K-means clustering, hierarchical cluster analysis (HCA), and a visual interpretation of VH polarization time series profiles are used to generate rice extent, cropping patterns, and spatiotemporal distribution of growth stages. To automate the process, four supervised classification methods (support vector machine (SVM), artificial neural networks (ANN), random forests, and C5.0 classification models) were independently trialled to identify cluster labels. The results from each classification method were compared. The method can also forecast rice extent for up to two months. The VH polarization data can identify four growth stages of rice—T&P: tillage and planting (30 days); V: vegetative-1 and 2 (60 days); R: reproductive (30 days); M: maturity (30 days). Compared to field survey data, this method measures overall rice extent with an accuracy of 96.5% and a kappa coefficient of 0.92. SVM and ANN show better performance than random forest and C5.0 models. This simple and robust method could be rolled out across Southeast Asia, and could be used as an alternative to time-consuming, expensive field surveys.

The utility of Sentinel-2 Vegetation Indices (VIs) and Sentinel-1 Synthetic Aperture Radar (SAR) for invasive alien species detection and mapping

The threat of invasive alien plant species is progressively becoming a serious global concern. Alien plant invasions adversely affect both ecological services and socio-economic systems. Hence, accurate detection and mapping of invasive alien species is valuable in mitigating adverse ecological and socio-economic effects. Recent advances in active and passive remote sensing technology have created new and cost-effective opportunities for the application of remote sensing to invasive species mapping. In this study, new generation Sentinel-2 (S2) optical imagery was compared to S2 derived Vegetation Indices (VIs) and S2 VIs fused with Sentinel-1 (S1) Synthetic Aperture Radar (SAR) imagery for detecting and mapping the American Bramble (Rubuscuneifolius). Fusion of S2 VIs and S1SAR imagery was conducted at pixel level and multi-class Support Vector Machine (SVM) image classification was used to determine the dominant land use land cover classes. Results indicated that S2 derived VIs were the most accurate (80%) in detecting and mapping Bramble, while fused S2 VIs and S1SAR were the least accurate (54%). Findings from this study suggest that the application of S2 VIs is more suitable for Bramble detection and mapping than the fused S2 VIs and S1SAR. The superior performance of S2 VIs highlights the value of the new generation S2 VIs for invasive alien species detection and mapping. Furthermore, this study recommends the use of freely available new generation satellite imagery for cost effective and timeous mapping of Bramble from surrounding native vegetation and other land use land cover types.

Divergent changes in cropping patterns and their effects on grain production under different agro-ecosystems over high latitudes in China

Drastic rice paddy expansion and rapid upland crop loss have occurred over high latitudes in China, which would affect national food security. Different agro-ecosystems (i.e., state farms guided by the central government for agriculture and private farms guided by individual farmers for agriculture) could lead to different agricultural land use patterns; but this topic has not been investigated, which has limited our understanding of the dynamics of cropping patterns (i.e., rice paddies and upland crops) under different agro-ecosystems and their effect on total grain production. Thus, this study examined these issues over high latitudes in China. The results showed that: the developed methodology for determining cropping patterns presented high accuracy (over 90%). Based on the cropping pattern data, first, a satellite evidence of substantial increase in rice paddies with the loss of upland crops was found, and the large-scale conversion from upland crops to rice paddies has become the principal land use changes during the period of 2000-2015. Second, the new phenomenon was observed with rice paddies in state farms expanding at faster rates (at proportions of 12.98%-70.11%) than those in private farms (4.86%-30.48%). Third, the conversion of upland crops into rice paddies contributed 10.69% of the net increase in grain, which played a significant role in ensuring food security. The study provided new evidence of different changes in cropping patterns under different agro-ecosystems, thereby affecting rice cropping pattern and total grain production. This information is important for understanding and guiding the response to food sustainability and environmental issues.

Efficient Identification of Corn Cultivation Area with Multitemporal Synthetic Aperture Radar and Optical Images in the Google Earth Engine Cloud Platform

The distribution of corn cultivation areas is crucial for ensuring food security, eradicating hunger, adjusting crop structures, and managing water resources. The emergence of high-resolution images, such as Sentinel-1 and Sentinel-2, enables the identification of corn at the field scale, and these images can be applied on a large scale with the support of cloud computing technology. Hebei Province is the major production area of corn in China, and faces serious groundwater overexploitation due to irrigation. Corn was mapped using multitemporal synthetic aperture radar (SAR) and optical images in the Google Earth Engine (GEE) cloud platform. A total of 1712 scenes of Sentinel-2 data and 206 scenes of Sentinel-1 data acquired from June to October 2017 were processed to composite image metrics as input to a random forest (RF) classifier. To avoid speckle noise in the classification results, the pixel-based classification result was integrated with the object segmentation boundary completed in eCognition software to generate an object-based corn map according to crop intensity. The results indicated that the approach using multitemporal SAR and optical images in the GEE cloud platform is reliable for corn mapping. The corn map had a high F1-Score of 90.08% and overall accuracy of 89.89% according to the test dataset, which was not involved in model training. The corn area estimated from optical and SAR images was well correlated with the census data, with an R2 = 0.91 and a root mean square error (RMSE) of 470.90 km2. The results of the corn map are expected to provide detailed information for optimizing crop structure and water management, which are critical issues in this region.

Reconstructing Geostationary Satellite Land Surface Temperature Imagery Based on a Multiscale Feature Connected Convolutional Neural Network

Geostationary satellite land surface temperature (GLST) data are important for various dynamic environmental and natural resource applications for terrestrial ecosystems. Due to clouds, shadows, and other atmospheric conditions, the derived LSTs are often missing a large number of values. Reconstructing the missing values is essential for improving the usability of the geostationary satellite LST data. However, current reconstruction methods mainly aim to fill the values of a small number of invalid pixels with many valid pixels, which can provide useful land surface temperature values. When the missing data extent becomes large, the reconstruction effect will worsen because the relationship between different spatiotemporal geostationary satellite LSTs is complex and highly nonlinear. Inspired by the superiority of the deep convolutional neural network (CNN) in solving highly nonlinear and dynamic problems, a multiscale feature connection CNN model is proposed to fill missing LSTs with large missing regions. The proposed method has been tested on both FengYun-2G and Meteosat Second Generation-Spinning Enhanced Visible and InfraRed Imager geostationary satellite LST datasets. The results of simulated and actual experiments show that the proposed method is accurate to within about 1 °C, with 70% missing data rates. This is feasible and effective for large regions of LST reconstruction tasks.

Classification of Rice Heavy Metal Stress Levels Based on Phenological Characteristics Using Remote Sensing Time-Series Images and Data Mining Algorithms

Heavy metal pollution in crops leads to phenological changes, which can be monitored by remote sensing technology. The present study aims to develop a method for accurately evaluating heavy metal stress in rice based on remote sensing phenology. First, the enhanced spatial and temporal adaptive reflectance fusion model (ESTARFM) was applied to blend Moderate Resolution Imaging Spectroradiometer (MODIS) and Landsat to generate a time series of fusion images at 30 m resolution, and then the vegetation indices (VIs) related to greenness and moisture content of the rice canopy were calculated to create the time-series of VIs. Second, phenological metrics were extracted from the time-series data of VIs, and a feature selection scheme was designed to acquire an optimal phenological metric subset. Finally, an ensemble model with optimal phenological metrics as classification features was built using random forest (RF) and gradient boosting (GB) classifiers, and the classification of stress levels was implemented. The results demonstrated that the overall accuracy of discrimination for different stress levels is greater than 98%. This study suggests that fusion images can be utilized to detect heavy metal stress in rice, and the proposed method may be applicable to classify stress levels.

Yield effect of applying earthworm castings produced during the oilseed rape-growing season in rice-oilseed rape cropping fields to rice

In-field earthworm density can be increased by planting oilseed rape during the non-rice growing season as compared to maintaining the rice-growing fields in fallow. This study was conducted to determine the effect on rice yield of earthworm castings produced during the oilseed rape-growing season in rice-oilseed rape cropping fields and to identify the critical factors that contribute to the yield effect. Field microplot experiments were conducted in 2016 and 2017. In 2016, a rice cultivar was grown under a factorial combination of absence (EC₀: 0 kg m⁻²) and presence of earthworm castings (EC₁: 17 kg m⁻²) with three N application rates (9, 12 and 15 g m⁻²). In 2017, nine rice cultivars were grown under EC₀ and EC₁ with the moderate N rate as was used in 2016. Results showed that application of earthworm castings produced during the oilseed rape-growing season in rice-oilseed rape cropping fields had a significant positive yield effect on rice. This was attributed to increased panicle m⁻² and total aboveground biomass while spikelets panicle⁻¹, spikelet filling percentage, grain weight, and harvest index were not affected. Our study indirectly provides a new evidence that oilseed rape is an excellent previous crop for cereals.