Article

Monitoring farmers' decisions on Mediterranean irrigated crops using satellite image time series

Taylor & Francis
International Journal of Remote Sensing
Authors:
  • Universitat Autònoma de Barcelona - Autonomous University of Barcelona
  • Universitat Autònoma de Barcelona, Catalonia, Spain
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

The main purpose of this study is to present a methodology for mapping and monitoring temporal signatures of Mediterranean crops over several years in irrigated areas, and to study their inter‐annual dynamics. These goals were achieved by remote sensing using 36 Landsat images from 2002 to 2005. Four crop maps, one for each year, with six agricultural categories and a thematic accuracy of 93%, 95%, 96% and 94% were obtained using a hybrid classifier. A mean of nine images produced these highly accurate results, but the absence of one image in the growth period of 2002 resulted in lower accuracies, particularly in fruit trees (85% of user accuracy). This highlights the importance of a multi‐temporal approach based on a relatively large number of images. After the classification results were validated, two parameters were used to characterize the dynamics of the four crops (rice, maize, alfalfa and fruit trees): greenness, extracted from the Normalized Difference Vegetation Index (NDVI), and wetness, calculated from the Tasselled Cap Wetness (TCW) Index. In order to differentiate the wetness origin of crops, an analysis of local daily precipitation (which could cause significant anomalies in the TCW coefficients) and water stored in the Susqueda reservoir (which may result in farmers making important management decisions when water is limited) was conducted during this four‐year period. After applying statistical analysis, the results showed that, of the four crops analysed, rice, alfalfa and fruit trees had more stable dynamics than maize, which was planted later in case of water deficit at the beginning of the irrigation campaign (in 2002) and earlier when the deficit occurred later (in 2005).

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... In order to identify and monitor different types of crops through satellite data, the normalized difference vegetation index (NDVI) has been one of the most commonly used indices because it is a powerful indicator of greenness, health, and growth of vegetation. It has been used alone or in combination with other indices such as the wetness component of the Tasselled Cap Transformation (Serra and Pons 2008;Shofiyati and Uchida 2011), and the normalized difference water index (Mulianga et al. 2015). It has been also used as input image in unsupervised classifications such as International Organization for Standardization. ...
... NDVI has been widely employed in multi-temporal approaches (e.g. Bradley et al. 2007;Wardlow and Egbert 2008;Serra and Pons 2008;Shofiyati and Uchida 2011;Zafar and Waqar 2014) because a single date image not always is sufficient to differentiate crops only on the basis of their spectral signatures (Jewell 1989;Van Niel and Mc Vicar 2004). Odenweller and Johnson (1984) firstly considered temporal-spectral profiles of vegetation indicators (Greenness above Bare Soil, GRABS, in their specific case) for crop identification purposes on Multi-Spectral Scanner Landsat images. ...
... Odenweller and Johnson (1984) firstly considered temporal-spectral profiles of vegetation indicators (Greenness above Bare Soil, GRABS, in their specific case) for crop identification purposes on Multi-Spectral Scanner Landsat images. Other studies followed, based on the use of crop-specific NDVI temporal profiles, for crop identification and mapping such as Murakami et al. (2001) on the high-resolution visible (HRV) and Navarro et al. (2016) on the high resolution geometric (HRG) sensor on board the Satellite Pour l'Observation de la Terre (SPOT), Guerschman et al. (2003) and Zafar and Waqar (2014) on TM Landsat, Ouzemou et al. (2015) on the Operational Land Imager (OLI) on Landsat 8. Serra and Pons (2008) and Foerster et al. (2012) also preferred a multi-temporal analysis of TM/ETM+ Landsat images for a refined crop type mapping. ...
Article
The use of remote sensing in the context of the Common Agricultural Policy (CAP) has progressively become an official method to support European (EU) Member States in carrying out controls about declarations of farmers requiring EU subsidies in agriculture. Reliable automatic or semi-automatic methodologies aiming at crop identification are still being developed and the only technique, which is officially accepted in the CAP context, remains photo interpretation of high/very high (satellite or aerial) orthoimages. To verify past situations, only orthophotos can be used but, unfortunately, they are not always available. In these cases, the use of satellite sensors with adequate spatial, spectral, and temporal resolutions, together with a reliable data analysis technique, could support or even substitute orthophoto interpretation. In this study, we propose a multi-temporal, multispectral algorithm exploiting the Thematic Mapper/Enhanced Thematic Mapper Plus data on Landsat platforms to identify different land covers in the context of CAP. Here it is presented to discriminate arable from non-arable lands. Assessment of the methodology was carried out using Corine 2012 and more than 1500 validation points over Basilicata region (Southern Italy). A general good agreement was found (74%), which increases to 82% in the specific case of arable land identification.
... For example, it was found that incorporation of crop phenology for crop classification can increase its accuracy. Serra and Pons (2008) [13] built a multitemporal model from 36 Landsat images (2002)(2003)(2004)(2005) to classify different Mediterranean crops (fruit, alfalfa, rice, winter cereals, maize, and other crops) [13]. Their model achieved over 93% accuracy when including multi-temporal signals. ...
... For example, it was found that incorporation of crop phenology for crop classification can increase its accuracy. Serra and Pons (2008) [13] built a multitemporal model from 36 Landsat images (2002)(2003)(2004)(2005) to classify different Mediterranean crops (fruit, alfalfa, rice, winter cereals, maize, and other crops) [13]. Their model achieved over 93% accuracy when including multi-temporal signals. ...
Article
Full-text available
A phenology-based crop type mapping approach was carried out to map cotton fields throughout the cotton-growing areas of eastern Australia. The workflow was implemented in the Google Earth Engine (GEE) platform, as it is time efficient and does not require processing in multiple platforms to complete the classification steps. A time series of Normalised Difference Vegetation Index (NDVI) imagery were generated from Landsat 8 Surface Reflectance Tier 1 (L8SR) and processed using Fourier transformation. This was used to produce the harmonised-NDVI (H-NDVI) from the original NDVI, and then phase and amplitude values were generated from the H-NDVI to visualise active cotton in the targeted fields. Random Forest (RF) models were built to classify cotton at early, mid and late growth stages to assess the ability of the model to classify cotton as the season progresses, with phase, amplitude and other individual bands as predictors. Results obtained from leave-one-season-out cross validation (LOSOCV) indicated that Overall Accuracy (OA), Kappa, Producer's Accuracies (PA) and User's Accuracy (UA), increased significantly when adding amplitude and phase as predictor variables to the model, than prediction using H-NDVI or raw bands only. Commission and omission errors were reduced significantly as the season progressed and more in-season imagery was available. The methodology proposed in this study can map cotton crops accurately based on the reconstruction of the unique cotton reflectance trajectory through time. This study confirms the importance of phenological metrics in improving in-season cotton fields mapping across eastern Australia. This model can be used in conjunction with other datasets to forecast yield based on the mapped crop type for improved decision making related to supply chain logistics and seasonal outlooks for production.
... Remote sensing (RS) has some advantages over traditional methods for mapping and monitoring vegetation status and the water balance with cost-effectiveness, temporal resolution and geographical extent being the most important attributes (Ambast et al., 2002;Incerti et al., 2007;Serra and Pons, 2008;Peña-Barragán et al., 2011). Crop water requirements have been obtained with RS data using different methods. ...
... According to the temporal NDVI profile (Figure 4(a)), artichokes and winter cereals showed the highest values in April and May and the lowest in August when the winter cereals had been harvested (Serra and Pons, 2008) and the artichokes, a perennial crop, remain dry as the downward trend indicates. In the case of maize, a summer cereal, the highest NDVI values occurred in June-July, when they are very green. ...
Article
The present study illustrates an original methodology for estimating irrigation requirements and quantifying real water consumption in a long-established Mediterranean rural community (Delta Llobregat, Barcelona, Spain), combining data from remote sensing, field mapping and in situ measurements. Because of land fragmentation and crop diversification, SPOT-5 imagery was used, given its spatial and temporal resolution and spectral attributes. Simultaneously, four flow meters were installed in two representative locations to measure water inputs and outputs every 5 min. Conveyance and irrigation efficiency were estimated for the entire irrigation community. The average conveyance efficiency was 46.8% and the classical and net irrigation efficiency reached 26.4 and 59.8%, respectively, with half of the water volume (55% or 3.2 hm3) returned to the river or diverted to wetlands, the maximum percentage of estimated error being about 3.4%. These results indicate an exceptionally high water loss rate due to the irrigation system (flooding), the ageing conveyance network and urban infrastructure breakdown. The applied protocol proved useful for monitoring low-efficiency irrigation systems in small communities experiencing intense urban and industrial pressures.
... Satellite imagery has become a promising and versatile source for global agricultural monitoring in recent years (Kussul et al., 2015(Kussul et al., , 2020Nandibewoor et al., 2015;Nguyen et al., 2020;Serra & Pons, 2008;Wu et al., 2015;Zhang et al., 2020). The use of image times series from the same scene at different periods has driven the interest in regularly monitoring crops at the Earth's surface (Valero et al., 2015). ...
Article
Full-text available
Purpose and Methods Crop identification using remotely sensed imagery provides useful information to make management decisions about land use and crop health. This research used phonecams to acquire the Normalized Difference Vegetation Index (NDVI) of various crops for three crop seasons. NDVI time series from Sentinel (L121-L192) images was also acquired using Google Earth Engine (GEE) for the same period. The resolution of satellite data is low therefore gap filling and smoothening filters were applied to the time series data. The comparison of data from satellite images and phenocam provides useful insight into crop phenology. The results show that NDVI is generally underestimated when compared to phenocam data. The Savitzky-Golay (SG) and some other gap filling and smoothening methods are applied to NDVI time series based on satellite images. The smoothened NDVI curves are statistically compared with daily NDVI series based on phenocam images as a reference. Results The SG method has performed better than other methods like moving average. Furthermore, polynomial order has been found to be the most sensitive parameter in applying SG filter in GEE. Sentinel (L121-L192) image was used to identify wheat during the year 2022–2023 in Sargodha district where experimental fields were located. The Random Forest Machine Leaning algorithm was used in GEE as a classifier. Conclusion The classification accuracy has been found 97% using this algorithm which suggests its usefulness in applying to other areas with similar agro-climatic characteristics.
... An early warning was developed based on a crucial phenological time. Phenology helps to obtain necessary information on plants, such as detecting, classifying, and monitoring plants [17]- [19], [28]- [30]. An early warning based on crucial phenology can anticipate problems with sugarcane crops and suggest practices to increase sugarcane productivity and production. ...
... This method is known as tasseled cap transformations (TCT) of multispectral scanner (MSS) and thematic mapper (TM) image data, which exemplify linear combination band features [29][30][31]. The TCT-induced three indices have been used extensively in satellite remote sensing studies of agriculture, landscape, ecology, and forest [32][33][34][35][36][37]. ...
Article
Full-text available
Bangladesh is a global south hotspot due to climate change and sea level rise concerns. It is a highly disaster-prone country in the world with active deltaic shorelines. The shorelines are quickly changing to coastal accretion and erosion. Erosion is one of the water hazards to landmass sinking, and accretion relates to land level rises due to sediment load deposition on the Bay of Bengal continental shelf. Therefore, this study aimed to explore shoreline status with change assessment for the three study years 1991, 2006, and 2021 using satellite remote sensing and geographical information system (GIS) approaches. Landsat 5, 7 ETM+, and 8 OLI satellite imageries were employed for onshore tasseled cap transformation (TCT) and land and sea classification calculations to create shore boundaries, baseline assessment, land accretion, erosion, point distance, and near feature analysis. We converted 16,550 baseline vertices to points as the study ground reference points (GRPs) and validated those points using the country datasheet collected from the Survey of Bangladesh (SoB). We observed that the delta’s shorelines were changed, and the overall lands were accredited for the land-increasing characteristics analysis. The total accredited lands in the coastal areas observed during the time periods from 1991 to 2006 were 825.15 km2, from 2006 to 2021 was 756.69 km2, and from 1991 to 2021 was 1223.94 km2 for the 30-year period. Similarly, coastal erosion assessment analysis indicated that the results gained for the period 1991 to 2006 and 2006 to 2021 were 475.87 km2 and 682.75 km2, respectively. Therefore, the total coastal erosion was 800.72 km2 from 1991 to 2021. Neat accretion was 73.94 km2 for the 30-year period from 1991 to 2021. This research indicates the changes in shorelines, referring to the evidence for the delta’s active formation through accretion and erosion processes of ‘climate change’ and ‘sea level rise’. This research projects the erosion process and threatens land use changes toward agriculture and settlements in the coastal regions of Bangladesh.
... In both of these cases, observed anomalies in growth and senescence or reductions in grain yield should also be considered against potential variations in climatic conditions season over season [33][34][35][36][37]. One other potential time series anomaly detection approach is to focus on rapid unusual changes, meaning to normalize the change rate of the NDVI curve through plotting the NDVI first derivative, which should help to remove the effects of climatic conditions, and other potential outside effects such as a lack of fertilizer or intercropping, which may vary between seasons [38][39][40][41][42]. ...
Article
Full-text available
The second United Nations Sustainable Development Goal (SDG2), zero hunger, aims to improve the productivity, food security, nutrition, and sustainability of small-scale farmers. The fall armyworm (FAW, Spodoptera frugiperda) has been devasting to smallholder farmer food security since it spread to sub-Saharan Africa in 2016, who have suffered massive crop losses, particularly maize, an important staple for basic sustenance. Since the FAW mainly devours green leaf biomass during the maize vegetative growth stage, the implementation of remote sensing technologies offers opportunities for monitoring the FAW. Here, we developed and tested a Sentinel 2 a+b satellite-based monitoring algorithm based on optimized first-derivative NDVI time series analysis using Google Earth Engine. For validation, we first employed the FAO Fall Armyworm Monitoring and Early Warning System (FAMEWS) mobile app data from Kenya, and then subsequently conducted field validation campaigns in Zimbabwe, Kenya, and Tanzania. Additionally, we directly observed loss of green biomass during maize vegetative growth stages caused by the FAW, confirming the observed signals of loss of the leaf area index (LAI) and the total green biomass (via the NDVI). Preliminary analyses suggested that satellite monitoring of small-scale farmer fields at the regional level may be possible with an NDVI first-derivative time series anomaly analysis using ESA Sentinel 2 a+b (R2 = 0.81). Commercial nanosatellite constellations, such as PlanetScope, were also explored, which may offer benefits from greater spatial resolution and return interval frequency. Due to other confounding factors, such as clouds, intercropping, weeds, abiotic stresses, or even other biotic pests (e.g., locusts), validation results were mixed. Still, maize biomass anomaly detection for monitoring the FAW using satellite data could help confirm the presence of the FAW with the help of expanded field-based monitoring through the FAO FAMEWS app.
... Mean shift segmentation was performed using ArcGIS Pro (ESRI, West Redlands, CA, USA) ( Figure 2C). In this platform, the spectral and spatial parameters are represented according to the level of importance (in the range of [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] given to the spectral differences and the importance given to the proximity, respectively. In this study, these parameters are referred to as spectral and spatial characteristics of pixels within an object. ...
Article
Full-text available
The overarching aim of this research was to develop a method for deriving crop maps from a time series of Sentinel-2 images between 2017 and 2018 to address global challenges in agriculture and food security. This study is the first step towards improving crop mapping based on phenological features retrieved from an object-based time series on a national scale. Five main crops in Israel were classified: wheat, barley, cotton, carrot, and chickpea. To optimize the object-based classification process, different characteristics and inputs of the mean shift segmentation algorithm were tested, including vegetation indices, three-band combinations, and high/low emphasis on the spatial and spectral characteristics. Four known vegetation indices (VIs)-based time series were tested. Additionally, we compared two widely used machine learning methods for crop classification, support vector machine (SVM) and random forest (RF), in addition to a newer classifier, extreme gradient boosting (XGBoost). Lastly, we examined two accuracy measures—overall accuracy (OA) and area under the curve (AUC)—in order to optimally estimate the accuracy in the case of imbalanced class representation. Mean shift best performed when emphasizing both the spectral and spatial characteristics while using the green, red, and near-infrared (NIR) bands as input. Both accuracy measures showed that RF and XGBoost classified different types of crops with significantly greater success than achieved by SVM. Nevertheless, AUC was better able to represent the significant differences between the classification algorithms than OA was. None of the VIs showed a significantly higher contribution to the classification. However, normalized difference infrared index (NDII) with XGBoost classifier showed the highest AUC results (88%). This study demonstrates that the short-wave infrared (SWIR) band with XGBoost improves crop type classification results. Furthermore, the study emphasizes the importance of addressing imbalanced classification datasets by using a proper accuracy measure. Since object-based classification and phenological features derived from a VI-based time series are widely used to produce crop maps, the current study is also relevant for operational agricultural management and informatics at large scales.
... Regional intra-class variation exists in a single agriculture crop type due to farmer's decisions to plant crops at different dates in different regions (Wardlow et al., 2007). These variations remain consistent within similar agriculture ecological zones (AEZs) and field size landscapes (e.g., large, medium, and small) due to similar agriculture and ecological conditions (Serra and Pons, 2008;Simonneaux et al., 2008). Therefore, the limited crop type reference data has the potential to be effectively extended by identifying the crop types in similar regions based on their spectral characteristics at specific growing stage/time (i.e., phenology). ...
Article
Full-text available
The combination of high spatial resolution and multi-date satellite imagery offers new opportunities for mapping and monitoring crop types of different agricultural field sizes. However, mapping of crop types at high spatial resolution requires high-quality crop type reference data typically collected from the ground-based surveys to create the maps and/or to assess the map accuracy. The availability of sufficient crop type reference data is limited over large geographic regions because of the time, effort, cost, and accessibility in different parts of the world. To generate large area crop type maps, any existing, but limited reference data must be spatially extended to other regions using appropriate and available non-ground-based sources. There is the potential to classify High Resolution Imagery (HRI) using a phenology-based approach as demonstrated in this paper to generate additional reference data within similar agriculture ecological zones (AEZs) based on the crop characteristics, their types, and their growing season. Therefore, the objective of this study was to evaluate if existing, limited crop type reference data could be extended using this approach. Multi-date, high spatial resolution satellite images were used to spatially extend the limited crop type reference data from one region [called the training region (TR)] to another region [called the test region (TE)] within the same AEZ using a phenology-based Decision Tree (DT) classifier for three different field sizes. The results demonstrate that this phenology-based classification approach can efficiently and effectively extend the limited crop type reference data to other regions in same AEZ for different field sizes.
... K-T transformation is performed on 20 remote sensing images, and three indexes are obtained for each remote sensing image, brightness, greenness and wetness. The brightness index is able to reflect the overall reflection effect of surface characteristics, the greenness index reflects the situation of surface vegetation, and the wetness index is able to reflect the water condition of the surface (Serra and Pons, 2008;Chen and Jiang, 2009). A corresponding extension tool in ENVI5.3, ...
Article
Full-text available
Remote sensing image data are often used as input in digital soil mapping (DSM). However, it is difficult to distinguish and identify soil types with less difference in reflectance spectral characteristics, because a small amount of input is not enough to provide enough common features. We consider that the hyper-temporal remote sensing data can be used to extract more common features of soil. The accuracy of DSM is improved by using the common features of soil or effective terrain attributes. We took Mingshui County of the Songnen Plain in northeast China as study area, which is known as a Black soil region. STRM DEM, legacy soil data, and 20 scenes Landsat images of bare soil period from 1984 to 2018 (April and May are considered a period of cultivated soil exposure in the study area), were used, with a maximum likelihood method classifier. A digital soil mapping model was constructed based on hyper-temporal data. Results from the study show that the accuracy of mapping with hyper-temporal classification characteristics, with an overall accuracy of 85.18% and a Kappa coefficient of 0.772, is higher than that of mono-temporal classification characteristics, with an average overall accuracy of 64.35% and an average Kappa coefficient of 0.467. After the introduction of relief degree of land surface (RDLS), the overall accuracy and Kappa coefficient of hyper-temporal mapping were 88.22% and 0.818, higher than the accuracy of other terrain factors. The research results signal the advantages of hyper-temporal remote sensing data in DSM, and the common features were able to improve the accuracy of DSM extracted from hyper-temporal data. This paper provided new insight to explain the impact of diverse terrain on DSM of Black soil region, and the mapping of soil type level could be accomplished more easily by the combination of the two characteristics.
... Also small field sizes require medium to highresolution images but these often lack temporal detail (Ozdogan, 2010). As a consequence, signals derived from remote sensing data often represent a mixture of crop types that require temporal 'un-mixing' or multi-temporal approaches for mapping multiple crops at the same time (Serra and Pons, 2008;Vyas et al., 2005;Zhong et al., 2015) but these approaches have not been applied on larger spatial scales yet. On the global scale, cropping or harvest intensity, the ratio of area harvested and physical area, has been mapped using national cropland data from the FAO (Ray and Foley, 2013) and gridded cropland data . ...
Article
Full-text available
Multiple cropping, defined as harvesting more than once a year, is a widespread land management strategy in tropical and subtropical agriculture. It is a way of intensifying agricultural production and diversifying the crop mix for economic and environmental benefits. Here we present the first global gridded data set of multiple cropping systems and quantify the physical area of more than 200 systems, the global multiple cropping area and the potential for increasing cropping intensity. We use national and sub-national data on monthly crop-specific growing areas around the year 2000 (1998–2002) for 26 crop groups, global cropland extent and crop harvested areas to identify sequential cropping systems of two or three crops with non-overlapping growing seasons. We find multiple cropping systems on 135 million hectares (12% of global cropland) with 85 million hectares in irrigated agriculture. 34%, 13% and 10% of the rice, wheat and maize area, respectively are under multiple cropping, demonstrating the importance of such cropping systems for cereal production. Harvesting currently single cropped areas a second time could increase global harvested areas by 87–395 million hectares, which is about 45% lower than previous estimates. Some scenarios of intensification indicate that it could be enough land to avoid expanding physical cropland into other land uses but attainable intensification will depend on the local context and the crop yields attainable in the second cycle and its related environmental costs.
... K-T transformation is performed on 20 remote sensing images, and three indexes are obtained for each remote sensing image, brightness, greenness and wetness. The brightness index is able to reflect the overall reflection effect of surface characteristics, the greenness index reflects the situation of surface vegetation, and the wetness index is able to reflect the water condition of the surface (Serra and Pons, 2008;Chen and Jiang, 2009). A corresponding extension tool in ENVI5.3, ...
Article
Remote sensing image data are often used as input in digital soil mapping (DSM). However, it is difficult to distinguish and identify soil types with less difference in reflectance spectral characteristics, because a small amount of input is not enough to provide enough common features. We consider that the hyper-temporal remote sensing data can be used to extract more common features of soil. The accuracy of DSM is improved by using the common features of soil or effective terrain attributes. We took Mingshui County of the Songnen Plain in northeast China as study area, which is known as a Black soil region. STRM DEM, legacy soil data, and 20 scenes Landsat images of bare soil period from 1984 to 2018 (April and May are considered a period of cultivated soil exposure in the study area), were used, with a maximum likelihood method classifier. A digital soil mapping model was constructed based on hyper-temporal data. Results from the study show that the accuracy of mapping with hyper-temporal classification characteristics, with an overall accuracy of 85.18% and a Kappa coefficient of 0.772, is higher than that of mono-temporal classification characteristics, with an average overall accuracy of 64.35% and an average Kappa coefficient of 0.467. After the introduction of relief degree of land surface (RDLS), the overall accuracy and Kappa coefficient of hyper-temporal mapping were 88.22% and 0.818, higher than the accuracy of other terrain factors. The research results signal the advantages of hyper-temporal remote sensing data in DSM, and the common features were able to improve the accuracy of DSM extracted from hyper-temporal data. This paper provided new insight to explain the impact of diverse terrain on DSM of Black soil region, and the mapping of soil type level could be accomplished more easily by the combination of the two characteristics.
... www.nature.com/scientificreports/ information 54 or plant-community degradation 55 or to supply information about crops 56 . Urban and water areas (NDVI ≤ 0) were masked from the original NDVI images to remove their signal from image texture analysis. ...
Article
Full-text available
Biodiversity monitoring at simultaneously fine spatial resolutions and large spatial extents is needed but limited by operational trade-offs and costs. Open-access data may be cost-effective to address those limitations. We test the use of open-access satellite imagery (NDVI texture variables) and biodiversity data, assembled from GBIF, to investigate the relative importance of variables of habitat extent and structure as indicators of bird community richness and dissimilarity in the Alentejo region (Portugal). Results show that, at the landscape scale, forest bird richness is better indicated by the availability of tree cover in the overall landscape than by the extent or structure of the forest habitats. Open-land birds also respond to landscape structure, namely to the spectral homogeneity and size of open-land patches and to the presence of perennial vegetation amid herbaceous habitats. Moreover, structure variables were more important than climate variables or geographic distance to explain community dissimilarity patterns at the regional scale. Overall, summer imagery, when perennial vegetation is more discernible, is particularly suited to inform indicators of forest and open-land bird community richness and dissimilarity, while spring imagery appears to be also useful to inform indicators of open-land bird richness.
... The other is to extract variables indicating temporal characteristics related to vegetation/crop using time series data of spectral reflectance values directly or derived vegetation index (Jakubauskas et al., 2002;Zhang et al., 2008;Sari et al., 2010;Conrad et al., 2011). The basis of this method is that the temporal domain of images holds much information on ecosystem behavior of vegetation/crops (Evans and Geerken, 2006), thus variables indicating temporal characteristics of crop can be used for crop identification with improved classification accuracy (Serra and Pons, 2008). Fourier transform has been used for detecting periodic patterns in time series data which are related to vegetation/crop features (Verhoef et al., 1996;Moody and Johnson, 2001;Geerken et al., 2005;Evans and Geerken, 2006;Zhang et al., 2008;Chen et al., 2018). ...
Article
Previous studies on soil organic carbon content or stock mapping mostly use natural environmental covariates and do not consider the soil management practice factor. However, human activities have become an important influencing factor for soil organic carbon, especially for agricultural soils. Crop species/crop rotations and management practices significantly affect the amount and spatial variation of soil organic carbon in croplands, but have not been considered for mapping soil organic carbon. In this study, we used direct crop rotation information and variables generated using Fourier transform on HJ-1A/1B NDVI time series data to capture the periodic effect of crop rotation, and explored the effectiveness of incorporating such information in predicting topsoil organic carbon content in cropland. A case study applied such method in a largely agricultural area in Anhui province, China. Crop rotation information was obtained through field investigation. Various combinations of predictive environmental variables were experimented for mapping soil organic carbon. The results were validated using field samples. Results showed that the combination of natural environment variables with both crop rotation type and variables derived through Fourier transform yielded the highest accuracy. In addition, only using the Fourier decomposed variables and crop rotation information were able to achieve a similar accuracy with using only soil formative natural environmental variables. This indicates that crop rotation information has comparable predictive power of soil organic carbon as natural environment variables. This study demonstrates the effectiveness of including agricultural practice information in digital soil mapping in agricultural landscapes with differences in crop rotation.
... In addition, the slopes of NDVI increase and NDVI decrease during the beginning and the ending of the season (i.e., NDVI Left Derivative and NDVI Right Derivative, respectively) might indicate the speeds at which the crop is growing after emergence and is ripening prior to harvest, respectively [10,33]. This information can also assist the decision-making processes of agrarian policy actions related to water supply and drought monitoring [50][51][52], food security [15], crop growing, and yield assessment [49], or the implementation of certain agro-environmental measures [53]. In addition, these remote-sensed products may be integrated with ecosystem models in order to assess greenhouse gas (GHG) cycling from agricultural crops at the regional scale and decrease the uncertainty in estimations of such models [54,55]. ...
Article
Full-text available
Remote sensing technology allows monitoring the progress of vegetation and crop phenology in large regions. Seasonal vegetation trends are commonly estimated from high temporal resolution but coarse spatial resolution satellite imagery, e.g., from MODIS-NDVI (Moderate Resolution Imaging Spectroradiometer—Normalized Difference Vegetation Index) time-series, which has usually limited their application to scenarios with few land uses or crops covering areas larger than actual parcel sizes. As an alternative, this paper proposes a general and robust procedure to map crop phenology at the level of individual crop parcels, and validates its feasibility in a complex and diverse cropland area located in central California. A first calibration phase consisted of evaluating the three curve-fitting models implemented in the TIMESAT software (i.e., asymmetric Gaussian (AG), double logistic (DL), and adaptive Savitzky–Golay (SG) filtering) and reporting the model and its settings that best adjusted to the MODIS-NDVI profile of each crop studied. Next, based on the selected crop-specific models and with a crop map previously obtained from ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) multi-temporal images, the procedure mapped four crop calendar events (i.e., start, end, middle, and length of the season) and five phenology-related metrics (i.e., base, maximum, amplitude, derivatives, and integrals of the NDVI values) of the study region by object-based image analysis (OBIA) of the MODIS-NDVI time-series. To mitigate the impact of mixed pixels, the OBIA procedure was designed to automatically apply a restrictive criterion based on the coverage of MODIS-NDVI pixels in each crop parcel: (1) using only the MODIS-NDVI pixels that were placed 100% within each crop parcel (i.e., “pure” pixels); or (2) if no “pure” pixels exist in any crop parcel, using only pixels with coverage percentages greater than 50%, and in such cases, reporting the mixing percentage in the output file. The calibration phase showed that the performance of the SG filtering was superior in most crops, with the exception of rice, while the AG model was intermediate in all of the cases. Differences between the dates of the start and end of the season that were observed in 120 ground-truth fields and the ones estimated by the crop-specific models were in a range of 11 days (for the corn fields) and 22 days (for the vineyard fields) on average. The OBIA procedure was also validated in 240 independent parcels with “pure” MODIS-NDVI pixels, reporting 89% and 82% of accuracy when mapping the start and end of the season, respectively. Our results revealed different growth patterns of the studied crops, especially of the crop calendar events of herbaceous (i.e., corn, rice, sunflower, and tomato) and woody crops (i.e., almond, walnut, and vineyard), of the NDVI derivatives of rice and the other studied herbaceous crops, and of the NDVI integrals of vineyard and the other studied woody crops. The resulting maps and tables provide valuable geospatial information for every parcel over time with several applications in cropland management, irrigation scheduling, and ecosystem modeling.
... NDVI values range from -1 (nonphotosynthetically active vegetation) to +1 (highly photosynthetically active vegetation). This index has been used successfully in several studies to evaluate land cover performance (Li et 190 al., 2004) or phenological information (Lloyd, 1990) or plant-community degradation (Alados et al., 2011) or to supply information about crops (Thenkabail et al., 1994;Martínez & Calera, 2001;Lyon et al., 2003;Jackson et al., 2004;Serra & Pons, 2008). Urban and water areas (NDVI ≤ 0) were masked from the original NDVI images to remove their signal from image texture analysis. ...
Preprint
Full-text available
Changes in ecosystem area are often used to assess human impacts on habitats and estimate biodiversity change. However, because species respond to structural changes at fine spatial scales the use of area alone may not capture all relevant changes. Operational costs limit the assessment of biodiversity change at a simultaneously fine spatial resolution and large scales. The development of cost-effective and expedite methods to monitor biodiversity change is therefore required. We use open access satellite imagery and biodiversity data to investigate the importance of variables of habitat extent and structure in explaining species richness and community dissimilarity of forest and open-land birds at the regional scale. Moreover, because Mediterranean landscapes are subject to seasonal dynamics, we explore the indicator value of remotely sensed variables measured in spring and summer. A large-scale dataset of bird occurrence data, including 8042 observations and 78 species, distributed by 40 landscape-sized cells, was assembled from GBIF after controlling for data quality. We found that summer satellite imagery, when the green perennial vegetation is more apparent, is particularly suited to model the diversity patterns of forest species, because distribution of tree cover in the landscape is well captured. Summer data is also useful to monitor the perennial elements that shape landscape structure and the habitat of open-land species. Specifically, mean NDVI and a second-order NDVI texture variable, were found to be good indicators of forest and open-land habitats, respectively. The use of spring imagery appears to be useful to monitor habitat structure within open-land habitat patches. Overall, NDVI texture measures were found to be good predictors of bird diversity patterns at large scales. Also, we were able to successfully conduct a regional scale analysis using open-access data, which illustrates their potential to inform large scale biodiversity monitoring.
... A matriz de erros, bem como as métricas de exatidão global e o índice Kappa, vem sendo utilizada para determinar a acurácia de classificações digitais mediante a utilização de imagens de satélites (Jupp, 1989;Congalton & Green, 1999;De Wit & Clevers, 2004;Serra & Pons, 2008;Peña-Barragán et al., 2011). Essa matriz permite avaliar o desempenho da classificação realizada para uma classe individual, particularmente quando um pequeno número de classes de uso do solo é de interesse, como, por exemplo, na estimativa de área de uma cultura agrícola (Ceballos-Silva & López-Blanco, 2003). ...
... A cropping pattern is a spatial and temporal arrangement of crops in rice fields (Manjunath et al., 2015). Cropping patterns result from the decision of farmers to optimize the use of resources (e.g., irrigation schedules, weather) (Serra and Pons, 2008). Cropping patterns possess spatial and temporal information that captures the responses of rice fields to a range of environmental and socioeconomic processes, including natural hazard occurrences (Xiao et al., 2005;Sakamoto et al., 2007;Boschetti et al., 2009;Sun et al., 2009;Gumma et al., 2011). ...
Article
Information on the vulnerability to flooding is vital to understand the potential damages from flood events. A method to determine the vulnerability to flooding in irrigated rice fields using the Enhanced Vegetation Index (EVI) was proposed in this study. In doing so, the time-series EVI derived from time-series 8 day 500 m spatial resolution MODIS imageries (MOD09A1) was used to generate cropping patterns in irrigated rice fields in West Java. Cropping patterns were derived from the spatial distribution and phenology metrics so that it is possible to show the variation of vulnerability in space and time. Vulnerability curves and cropping patterns were used to determine the vulnerability to flooding in irrigated rice fields. Cropping patterns capture the shift in the vulnerability, which may lead to either an increase or decrease of the degree of damage in rice fields of origin and other rice fields. The comparison of rice field areas between MOD09A1 and ALOS PALSAR and MOD09A1 and Agricultural Statistics showed consistent results with R² = 0.81 and R² = 0.93, respectively. The estimated and observed DOYs showed RMSEs = 9.21, 9.29, and 9.69 days for the Start of Season (SOS), heading stage, and End of Season (EOS), respectively. Using the method, one can estimate the relative damage provided available information on the flood depth and velocity. The results of the study may support the efforts to reduce the potential damages from flooding in irrigated rice fields.
... Odenweller & Johnson (1984) assessed multi-temporal Landsat images in the US Corn Belt and analysed the spectral profile of a green vegetation indicator of specific annual crops, such as winter cereal and sunflowers. Serra & Pons (2008) supported the use of a multi-temporal approach as well as phenology data in any crop classification methodology. The method used by Zhong et al. (2014) was based on spectral and phenological metric indices representing multi-temporal information with only a single year training set. ...
Article
Full-text available
Remote sensing (RS) offers an efficient and reliable means to map features on Earth. Crop type mapping using RS at various temporal and spatial resolutions plays an important role spanning from environmental to economical. The main objective of the current study was to evaluate the significance of optical data in a multi-temporal crop type classification-based on very high spatial resolution and high spatial resolution imagery. With this aim, three images from WorldView-3 and Sentinel-2 were acquired over Coalville (UK) between April and July 2016. Three vegetation indices (VIs); the normalized difference vegetation index, the green normalized difference vegetation index and soil adjusted vegetation index were generated using red, green and near-infrared spectral bands; then a supervised classification was performed using ground reference data collected from field surveys, Random forest (RF) and decision tree (DT) classification algorithms. Accuracy assessment was undertaken by comparing the classified output with the reference data. An overall accuracy of 91% and κ coefficient of 0·90 were estimated using the combination of RF and DT classification algorithms. Therefore, it can be concluded that integrating very high- and high-resolution imagery with different VIs can be implemented effectively to produce large-scale crop maps even with a limited temporal-dataset.
... These three indices have been widely studied and successfully used in studies of agriculture, forest, ecology, and landscape. For instance, the wetness component was used to characterize the dynamics of irrigated crops (Serra and Pons 2008). In recent years, the Tasseled Cap transformation was used as an effective method in wetland extraction via remote sensing technology. ...
Article
Full-text available
In this study, the spatial and temporal impacts of the Atatürk Dam on agro-meteorological aspects of the Southeastern Anatolia region have been investigated. Change detection and environmental impacts due to water-reserve changes in Atatürk Dam Lake have been determined and evaluated using multi-temporal Landsat satellite imageries and meteorological datasets within a period of 1984–2011. These time series have been evaluated for three time periods. Dam construction period constitutes the first part of the study. Land cover/use changes especially on agricultural fields under the Atatürk Dam Lake and its vicinity have been identified between the periods of 1984–1992. The second period comprises the 10-year period after the completion of filling up the reservoir in 1992. At this period, Landsat and meteorological time-series analyses are examined to assess the impact of the Atatürk Dam Lake on selected irrigated agricultural areas. For the last 9-year period from 2002 to 2011, the relationships between seasonal water-reserve changes and irrigated plains under changing climatic factors primarily driving vegetation activity (monthly, seasonal, and annual fluctuations of rainfall rate, air temperature, humidity) on the watershed have been investigated using a 30-year meteorological time series. The results showed that approximately 368 km2 of agricultural fields have been affected because of inundation due to the Atatürk Dam Lake. However, irrigated agricultural fields have been increased by 56.3% of the total area (1552 of 2756 km2) on Harran Plain within the period of 1984–2011.
... O sensoriamento remoto e os métodos tradicionais de classificação digital de imagens e avaliação de sua acurácia já são utilizados em alguns países (Gallego, 2004;Gallego et al., 2008;Wu et al., 2012). A matriz de erros, bem como as métricas de exatidão global e o índice Kappa, vem sendo utilizada para determinar a acurácia de classificações digitais mediante a utilização de imagens de satélites (Jupp, 1989;Congalton & Green, 1999;De Wit & Clevers, 2004;Serra & Pons, 2008;Peña‑Barragán et al., 2011). Essa matriz permite avaliar o desempenho da classificação realizada para uma classe individual, particularmente quando um pequeno número de classes de uso do solo é de interesse, como, por exemplo, na estimativa de área de uma cultura agrícola (Ceballos‑Silva & López‑Blanco, 2003).Chen & Goodchild (2007)também desenvolveram um método de calibração de estimativa de área de culturas agrícolas ao utilizar proporções da matriz de erros. ...
Article
Resumo – O objetivo deste trabalho foi estimar a área plantada com soja por meio da normalização da matriz de erros gerada a partir da classificação supervisionada de imagens TM/Landsat‑5. Foram avaliados oito municípios no Estado do Paraná, com dados referentes à safra de 2003/2004. As classificações foram realizadas por meio dos métodos paralelepípedo e máxima verossimilhança, dando origem à " máscara de soja ". Os valores do índice Kappa dos oito municípios ficaram acima de 0,6. As estimativas de área de soja, corrigidas por matriz de erros, apresentaram alta correlação com as estimativas oficiais do estado e com as estimativas geradas a partir de um método alternativo denominado " expansão direta ". A estimativa de área de soja por meio da normalização da matriz de erros apresenta menor custo e pode subsidiar métodos convencionais na estimativa menos subjetiva de safras. Termos para indexação: Glycine max, cultura da soja, geotecnologia, índice Kappa, previsão de safras, TM/Landsat‑5. Soybean crop area estimation through image classification normalized by the error matrix Abstract – The objective of this work was to estimate soybean crop area by the normalization of the error matrix generated from the supervised classification of TM/Landsat‑5 images. Eight municipalities of the state of Paraná, Brazil, were evaluated using data from the 2003/2004 crop season. Classifications were carried out using the parallelepiped and maximum likelihood methods, resulting in a " soybean mask ". Kappa index values for the eight municipalities were above 0.6. Estimated soybean areas, corrected by the error matrix, were highly correlated with official estimates of the state and with estimates generated from an alternative method called " direct expansion ". Soybean crop area estimation by the normalization of the error matrix is less costly and can aid conventional methods in estimating harvests in a less subjective manner.
... SITS offers opportunities for understanding how a location is changing, and, in some cases, also predicting the future changes. Therefore, it is a fundamental tool for environmental monitoring and analysis of land-cover dynamics, such as for monitoring the land use by human activities [14][15][16][17]. In some previous papers, for instance, we have shown examples of SITS to investigate the motion of sand dunes [18][19][20][21][22]. ...
... Instead, the Orchard class shows the greater commission (6-8% range) and omission (5-10% range) errors related to Herbaceous class, and vice versa with a slightly higher percentages. One of the encountered causes is the confusion between young orchard trees and the ripe herbaceous plants [51], [52]. The Bare soil class is mainly confused with Orchard and Herbaceous classes, mostly in the classifications of 1984 and 2015. ...
... With such spatially coarse data, a multi-temporal approach is considered essential to obtain high accuracy and the incorporation of crop phenology is recommended in the classification. It is also the case even for medium spatial resolution data such as Landsat with 30 m resolution (Oetter et al., 2001;De Wit and Clevers, 2004;Martínez-Casasnovas et al., 2005;Duchemin et al., 2006;Serra and Pons, 2008;Simonneaux et al., 2008;Cerqueira Leite et al., 2011;Vieira et al., 2012;Zhong et al., 2013), SPOT with 20 or 10 m resolution (Murakami et al., 2001;Yang et al., 2011;Duro et al., 2012), ASTER with 15 or 30 m resolution (Peña-Barragan et al., 2011), IRS LISS III radar imagery with 24 m-resolution (Mathur and Foody, 2008;Heller et al., 2012), SAR imagery (Tso and Mather, 1999;Vicente-Guijalba et al., 2014) or even combining AVHRR or MODIS with Landsat imagery (Fisher and Mustard, 2007;Velpuri et al., 2009), Landsat with airborne radar (Ulaby et al., 1982), SPOT/Landsat with ENVISAT/SAR/ASAR imagery (Laurila et al., 2010), Landsat with ENVISAT/MERIS (Amorós-López et al., 2013). More recently, some authors have used RapidEye with high spatial resolution (6.5 m) or even Formosat-2 data with both high spatial (8 m) and temporal (daily revisit) resolutions and constant viewing angles (Courault et al., 2008;Bsaibes et al., 2009;Claverie et al., 2012) and demonstrated evidence of their usefulness for crop mapping and/or monitoring. ...
Article
This study is part of several projects aiming at spatially monitoring the effects of exogeno us organic matter apply on soil organic carbon sequestration, and necessitating for this purpose the gathering of spatial data about cropping systems. The aim of this study was to assess the contribution of very high spatial resolution (VHR) Pléiades images to both early season crop identification and mapping and changes in bare soil surface characteristics due to cultural operations. The study region covering 4000 ha including 2100 ha-croplands is located west of the peri-urban territory of the Versailles plain and the Alluets plateau (Yvelines, France). About 100 cropped fields were observed on the ground synchronously with two Pléiades images of 3 and 24 April 2013 and one SPOT4 image of 2 April 2013. The GIS structuring of these field data was used for delimitating both training and test zones for the support vector machine classifier with polynomial function kernel (pSVM). For the single date-images, the pSVM was computed on the 4 spectral bands while for the bitemporal Pléiades image, it relied on the 8 spectral bands and the two NDVI bands. For the single-date classifications of crops, the overall accuracy reached 87[%] for the SPOT4 image of 2 April (6 classes), 79[%] for the Pleiades image of 3 April (6 classes) and 82[%] for that of 24 April (7 classes). For the bi-temporal Pléiades image, the overall accuracy was about 80[%] (7 classes), winter crops, grasslands and fallows being very well detected while confusions occured between spring barley at initial stages (2-3 leaves) and bare soils prepared for other spring crops. At the earlier date (2-3 April), the Pléiades image very well discriminated cultural operations (>77[%], user's or producer's accuracies) as well as fallows and grasslands, and brought unique information about within-field spatial heterogeneity of crop development stages, while winter cereals and rapeseed were better discriminated by the SPOT4 image winter cereals (>70[%], user's or producer's accuracies). Pleiades images therefore bring information complementary to multispectral images with high spatial resolution.
... TCT를 활용한 연구로는 작황의 변화 특성에 대한 연 구 (Serra and Pons, 2008), 식생의 타입과 수확 정도를 알 아내는 연구 (Jin and Sader, 2005b), 식생의 나이를 추정하 는 연구 (Wulder et al., 2004), 식생수분도를 추정하는 연 구 (Pereira, 2006) 및 육상의 식생 분류 연구 (Oetter et al., 2001;Dymond et al., 2002) 등이 있다. 이후 TCT의 정성 적인 분석의 한계를 극복하기 위해 육상 파라미터와 비 교하여 정량적인 결과를 산출하기 위한 시도가 있었다. ...
Article
The objective of this study is to determine Tasseled Cap Transformation (TCT) coefficients for the Geostationary Ocean Color Imager (GOCI). TCT is traditional method of analyzing the characteristics of the land area from multi spectral sensor data. TCT coefficients for a new sensor must be estimated individually because of different sensor characteristics of each sensor. Although the primary objective of the GOCI is for ocean color study, one half of the scene covers land area with typical land observing channels in Visible-Near InfraRed (VNIR). The GOCI has a unique capability to acquire eight scenes per day. This advantage of high temporal resolution can be utilized for detecting daily variation of land surface. The GOCI TCT offers a great potential for application in near-real time analysis and interpretation of land cover characteristics. TCT generally represents information of "Brightness", "Greenness" and "Wetness". However, in the case of the GOCI is not able to provide "Wetness" due to lack of ShortWave InfraRed (SWIR) band. To maximize the utilization of high temporal resolution, "Wetness" should be provided. In order to obtain "Wetness", the linear regression method was used to align the GOCI Principal Component Analysis (PCA) space with the MODIS TCT space. The GOCI TCT coefficients obtained by this method have different values according to observation time due to the characteristics of geostationary earth orbit. To examine these differences, the correlation between the GOCI TCT and the MODIS TCT were compared. As a result, while the GOCI TCT coefficients of "Brightness" and "Greenness" were selected at 4h, the GOCI TCT coefficient of "Wetness" was selected at 2h. To assess the adequacy of the resulting GOCI TCT coefficients, the GOCI TCT data were compared to the MODIS TCT image and several land parameters. The land cover classification of the GOCI TCT image was expressed more precisely than the MODIS TCT image. The distribution of land cover classification of the GOCI TCT space showed meaningful results. Also, "Brightness", "Greenness", and "Wetness" of the GOCI TCT data showed a relatively high correlation with Albedo (R^2
... The Landsat imagery with a 30 m spatial resolution was also found well suited for crop classification. For example, thirty-six Landsat images obtained from 2002 to 2005 were applied to extract temporal signatures of six main Mediterranean crops [8]. Time series of eight Landsat images were used to identify four main classes (bare soils, annual vegetation, trees on bare soil, and trees on annual understory) [9]. ...
Article
Full-text available
Cropland mapping via remote sensing can provide crucial information for agri-ecological studies. Time series of remote sensing imagery is particularly useful for agricultural land classification. This study investigated the synergistic use of feature selection, Object-Based Image Analysis (OBIA) segmentation and decision tree classification for cropland mapping using a finer temporal-resolution Landsat-MODIS Enhanced time series in 2007. The enhanced time series extracted 26 layers of Normalized Difference Vegetation Index (NDVI) and five NDVI Time Series Indices (TSI) in a subset of agricultural land of Southwest Missouri. A feature selection procedure using the Stepwise Discriminant Analysis (SDA) was performed, and 10 optimal features were selected as input data for OBIA segmentation, with an optimal scale parameter obtained by quantification assessment of topological and geometric object differences. Using the segmented metrics in a decision tree classifier, an overall classification accuracy of 90.87% was achieved. Our study highlights the advantage of OBIA segmentation and classification in reducing noise from in-field heterogeneity and spectral variation. The crop classification map produced at 30 m resolution provides spatial distributions of annual and perennial crops, which are valuable for agricultural monitoring and environmental assessment studies.
... With such spatially coarse data, a multi-temporal approach is considered essential to obtain high accuracy and the incorporation of crop phenology is recommended in the classification. It is also the case even for medium spatial resolution data such as Landsat with 30 m resolution (Oetter et al., 2001;De Wit and Clevers, 2004;Martínez-Casasnovas et al., 2005;Duchemin et al., 2006;Serra and Pons, 2008;Simonneaux et al., 2008;Cerqueira Leite et al., 2011;Vieira et al., 2012;Zhong et al., 2013), SPOT with 20 or 10 m resolution (Murakami et al., 2001;Yang et al., 2011;Duro et al., 2012), ASTER with 15 or 30 m resolution (Peña-Barragan et al., 2011), IRS LISS III radar imagery with 24 m-resolution (Mathur and Foody, 2008;Heller et al., 2012), SAR imagery (Tso and Mather, 1999;Vicente-Guijalba et al., 2014) or even combining AVHRR or MODIS with Landsat imagery (Fisher and Mustard, 2007;Velpuri et al., 2009), Landsat with airborne radar (Ulaby et al., 1982), SPOT/Landsat with ENVISAT/SAR/ASAR imagery (Laurila et al., 2010), Landsat with ENVISAT/MERIS (Amorós-López et al., 2013). More recently, some authors have used RapidEye with high spatial resolution (6.5 m) or even Formosat-2 data with both high spatial (8 m) and temporal (daily revisit) resolutions and constant viewing angles (Courault et al., 2008;Bsaibes et al., 2009;Claverie et al., 2012) and demonstrated evidence of their usefulness for crop mapping and/or monitoring. ...
Article
The aim of this study was to assess the contribution of very high spatial resolution (VHSR) Pléiades images to both early season crop identification and the mapping of bare soil surface characteristics due to cultural operations. The study region covering 21 km2 is located west of the peri-urban territory of the Versailles plain and the Alluets plateau (Yvelines, France). About 100 cropped fields were observed on the ground synchronously with two Pléiades images of 3 and 24 April 2013 and one SPOT4 image of 2 April 2013. The GIS structuring of these field data along with vector information about field boundaries was used for delimitating both training and test zones for the support vector machine classifier with polynomial function kernel (pSVM). The pSVM was computed on the spectral bands and NDVI for both single-date Pléiades and the bi-temporal Pléiades pair. For the single-date classifications of crops, the overall per-pixel accuracy reached 87% for the SPOT4 image of 2 April (6 classes), 79% for the Pléiadesimage of 3 April (6 classes) and 82% for that of 24 April (7 classes). At the earlier date (2–3 April), the Pléiades image very well discriminated cultural operations (>77%, user’s or producer’s accuracies) as well as fallows and grasslands, while winter cereals and rapeseed were better discriminated by the SPOT4 image (>70%, user’s or producer’s accuracies). As Pléiades images revealed within-field spatial variations of early phenological stages of winter cereals that could be critical for adjusting management of zones with delayed development during the growing season, they brought information complementary to multispectral images with high spatial resolution. For the bi-temporal Pléiades image,the overall per-pixel accuracy was about 80% (7 classes), winter crops, grasslands and fallows being very well detected while confusions occurred between spring barley at initial stages (2–3 leaves) and bare soils prepared for other spring crops. Using an additional validation field set covering ∼1/3 of the study area croplands, the crop map resulting from the bi-temporal Pléiades pair achieved correct crop prediction for about 89.7% of the validation fields when considering composite classes for winter cereals and for spring crops. Early-season Pléiades images therefore show a considerable potential for anticipating regional crop patterns and detecting soil tillage operations in spring.
... Various investigations have demonstrated the benefits of crop mapping using remote sensing data (Congalton et al., 1998;Oetter et al., 2001;Ulaby et al., 1982). Utilisation of time series satellite data was proved to be essential for high accuracy of crop classification (Barbosa et al., 1996;Serra and Pons, 2008;Simonneaux et al., 2008). ...
... The identification of outdoor crops by using Landsat time series has been already addressed by many authors [22,23]. Recently, an innovative methodology, which combines multi-temporal moderate resolution remote sensing data, the OBIA approach and the decision tree (DT) classifier algorithm, has been proposed [24,25]. ...
Article
Full-text available
Greenhouse detection and mapping via remote sensing is a complex task, which has already been addressed in numerous studies. In this research, the innovative goal relies on the identification of greenhouse horticultural crops that were growing under plastic coverings on 30 September 2013. To this end, object-based image analysis (OBIA) and a decision tree classifier (DT) were applied to a set consisting of eight Landsat 8 OLI images collected from May to November 2013. Moreover, a single WorldView-2 satellite image acquired on 30 September 2013, was also used as a data source. In this approach, basic spectral information, textural features and several vegetation indices (VIs) derived from Landsat 8 and WorldView-2 multi-temporal satellite data were computed on previously segmented image objects in order to identify four of the most popular autumn crops cultivated under greenhouse in Almería, Spain (i.e., tomato, pepper, cucumber and aubergine). The best classification accuracy (81.3% overall accuracy) was achieved by using the full set of Landsat 8 time series. These results were considered good in the case of tomato and pepper crops, being significantly worse for cucumber and aubergine. These OPEN ACCESS Remote Sens. 2015, 7 7379 results were hardly improved by adding the information of the WorldView-2 image. The most important information for correct classification of different crops under greenhouses was related to the greenhouse management practices and not the spectral properties of the crops themselves.
... Various investigations have demonstrated the benefits of crop mapping using remote sensing data (Congalton et al., 1998;Oetter et al., 2001;Ulaby et al., 1982). Utilisation of time series satellite data was proved to be essential for high accuracy of crop classification (Barbosa et al., 1996;Serra and Pons, 2008;Simonneaux et al., 2008). ...
Article
Full-text available
Crop mapping and time series analysis of agronomic cycles are critical for monitoring land use and land management practices, and analysing the issues of agro-environmental impacts and climate change. Multi-temporal Landsat data can be used to analyse decadal changes in cropping patterns at field level, owing to its medium spatial resolution and historical availability. This study attempts to develop robust remote sensing techniques, applicable across a large geographic extent, for state-wide mapping of cropping history in Queensland, Australia. In this context, traditional pixel-based classification was analysed in comparison with image object-based classification using advanced supervised machine-learning algorithms such as Support Vector Machine (SVM). For the Darling Downs region of southern Queensland we gathered a set of Landsat TM images from the 2010–2011 cropping season. Landsat data, along with the vegetation index images, were subjected to multiresolution segmentation to obtain polygon objects. Object-based methods enabled the analysis of aggregated sets of pixels, and exploited shape-related and textural variation, as well as spectral characteristics. SVM models were chosen after examining three shape-based parameters, twenty-three textural parameters and ten spectral parameters of the objects. We found that the object-based methods were superior to the pixel-based methods for classifying 4 major landuse/land cover classes, considering the complexities of within field spectral heterogeneity and spectral mixing. Comparative analysis clearly revealed that higher overall classification accuracy (95%) was observed in the object-based SVM compared with that of traditional pixel-based classification (89%) using maximum likelihood classifier (MLC). Object-based classification also resulted speckle-free images. Further, object-based SVM models were used to classify different broadacre crop types for summer and winter seasons. The influence of different shape, textural and spectral variables, and their weights on crop-mapping accuracy, was also examined. Temporal change in the spectral characteristics, specifically through vegetation indices derived from multi-temporal Landsat data, was found to be the most critical information that affects the accuracy of classification. However, use of these variables was constrained by the data availability and cloud cover.
... Although in other areas of US such as the Central Great Plain multitemporal AVHRR and MODIS images were successfully applied in mapping specific crop types (Wardlow et al., 2007;Ozdogan and Gutman, 2008), the coarse resolution is unlikely to effectively map crop fields in California where parcels are relatively small (Wardlow et al., 2007;Jakubauskas et al., 2001). Medium resolution images from the Landsat TM/ETMϩ (Thematic Mapper / Enhanced Thematic Mapper Plus) and SPOT (Satellite Pour l'Observation de la Terre) have proven suitable for discriminating crops as well as retrieving land parcel at a finer scale (Xie et al., 2007;Erol and Akdeniz, 2005;Martinez-Casasnovas et al., 2005;Murakami et al., 2001;Turker and Arikan, 2005;Serra and Pons, 2008;Conrad et al., 2010). The cropland data layer (CDL) of USDA National Agricultural Statistics Service (NASS) is a detailed, state-level crop classification. ...
Article
The overarching goal of this study was to map specific crop types in the Central Valley, California and estimate the effect of classification uncertainty on the calculation of crop evapotranspiration (ETc). A phenology-based classification (PBC) approach was developed to identify crop types based on phenological and spectral metrics derived from the time series of Landsat TM/ETM+ imagery. Phenological metrics, calculated by fitting asymmetric double sigmoid functions to temporal profiles of enhanced vegetation index (EVI), were capable of separating crop types with distinct crop calendars. An innovative method was used to compute spectral metrics to represent crops’ spectral characteristics at certain phenological stages instead of any specific imaging date. Crop mapping using these metrics showed a stable performance without influences of low-quality data and inter-annual differences in imaging dates. The requirement for ground reference data by the PBC approach was low because classification algorithms were mostly built according to the knowledge on crop calendars and agricultural practices. Techniques including image segmentation, data fusion with MODIS imagery, and decision tree were incorporated to make the approach effective and efficient. Though moderate accuracy (ε65.0 percent) was achieved, ETc calculated by the Food and Agriculture Organization (FAO) 56 method showed that the estimate of water use was not likely to be significantly affected by the classification error in PBC. All these advantages imply the strength of the PBC approach in the regular crop mapping of the Central Valley.
... The images were recorded on 16 May, 01 and 17 June, 19 July, 04 August, 23 October, and 08 November. According to some agricultural studies and our field experience (Serra et al., 2003;Serra and Pons, 2008), the main herbaceous and permanent crops cultivated in the study area were: dry winter cereals, irrigated maize, irrigated alfalfa, irrigated rice, dry and irrigated fruit trees, fallow land, dry olive trees, dry vineyards, other dry and irrigated herbaceous crops, and pastures. Figure 2 summarizes the methodology applied in this work. ...
Article
Full-text available
In this paper, cadastre agricultural cartography was enriched using crop raster maps obtained from remote sensing images. The work demonstrates the implications of applying two new terms: fidelity and purity. Per-pixel classifications and polygon enrichments were compared taking into account: (a) the consequences of using a more or less conservative strategy at the classification stage, using fidelity, and (b) the consequences of using modal thresholds at the enrichment stage when deciding which category each polygon is to be assigned to, using purity. More than 300,000 pixels and 2,800 polygons were used to measure the thematic accuracy of ten agricultural categories by means of confusion matrices. These were computed at pixel, polygon, and area level. Thematic accuracy was calculated in the classical way and without taking into account unclassified pixels as errors, as well as by paying special attention to the consequences for the classified area. The results show that polygon enrichment is a useful methodology, achieving thematic accuracies of 95.6 percent, when optimum parameters are used, while classifying 87.4 percent of the area.
... For example, Modis imagery of 250 m of pixel can only be appropriated for parcels that are larger than 32 ha [12,13]. Serra and Pons (2008;[14]) presented a methodology for mapping and monitoring the temporal signatures of six main Mediterranean crops (winter wheat, rice, corn, alfalfa, fruit trees and others) in central Spain using a hybrid classifier and Landsat images, reported that a multitemporal approach is essential and recommended the incorporation of phenology data in the classification methodology. Simoneaux et al. (2008) [15] used a time series of eight Landsat images over Morocco to identify four main classes (bare soil, annual vegetation, trees on bare soil and trees on annual understory), although these authors concluded that a precise typology of the crops could not be obtained based on the Normalized Difference Vegetation Index [NDVI = (NIR-R)/(NIR+R)] profiles. ...
Article
Full-text available
A procedure named CROPCLASS was developed to semi-automate census parcel crop assessment in any agricultural area using multitemporal remote images. For each area, CROPCLASS consists of a) a definition of census parcels through vector files in all of the images; b) the extraction of spectral bands (SB) and key vegetation index (VI) average values for each parcel and image; c) the conformation of a matrix data (MD) of the extracted information; d) the classification of MD decision trees (DT) and Structured Query Language (SQL) crop predictive model definition also based on preliminary land-use ground-truth work in a reduced number of parcels; and e) the implementation of predictive models to classify unidentified parcels land uses. The software named CROPCLASS-2.0 was developed to semi-automatically perform the described procedure in an economically feasible manner. The CROPCLASS methodology was validated using seven GeoEye-1 satellite images that were taken over the LaVentilla area (Southern Spain) from April to October 2010 at 3- to 4-week intervals. The studied region was visited every 3 weeks, identifying 12 crops and others land uses in 311 parcels. The DT training models for each cropping system were assessed at a 95% to 100% overall accuracy (OA) for each crop within its corresponding cropping systems. The DT training models that were used to directly identify the individual crops were assessed with 80.7% OA, with a user accuracy of approximately 80% or higher for most crops. Generally, the DT model accuracy was similar using the seven images that were taken at approximately one-month intervals or a set of three images that were taken during early spring, summer and autumn, or set of two images that were taken at about 2 to 3 months interval. The classification of the unidentified parcels for the individual crops was achieved with an OA of 79.5%.
... One of the most fundamental reasons for this problem is the variability within the same crop because of the variations in crop development schedule due to the farmer's decisions, local weather conditions, and other factors. Therefore, multi-temporal as well as seasonal image data acquisition representing crop phenology is essential to achieve high classification accuracy (Serra and Pons 2008;Peña-Barragán et al. 2011). Our study supports this conclusion by demonstrating that the two-season multi-spectral information improved the classification results, showing the highest overall accuracy and the highest certainty compared with other classification scenarios employing one-season image data. ...
Article
Full-text available
Recent satellite missions have provided new perspectives by offering high spatial resolution, a variety of spectral properties, and fast revisit rates to the same regions. In this study, we examined the utility of both broadband red-edge spectral information and texture features for classifying paddy rice crops in South Korea into three different growth stages. The rice grown in South Korea can be grouped into early-maturing, medium-maturing, and medium-late-maturing cultivars, and each cultivar is known to have a minimum and maximum productivity. Therefore, the accurate classification of paddy rice crops into a certain time line enables pre-estimation of the expected rice yields. For the analysis, two seasons of RapidEye satellite image data were used. The results showed that the broadband red-edge information slightly improved the classification accuracy of the paddy rice crops, particularly when single-season image data were used. In contrast, texture information resulted in only minor improvement or even a slight decline in accuracy, although it is known to be advantageous for object-based classification. This was due to the homogeneous nature of paddy rice fields, as different rice cultivars are similar in terms of their morphology. Based on these results, we conclude that the additional spectral information such as the red-edge band is more useful than the texture features to detect different crop conditions in relatively homogeneous rice paddy environments. We therefore confirm the potential of broadband red-edge information to improve the classification of paddy rice crops. However, there is still a need to examine the relationship between textural properties and paddy rice crop parameters in greater depth.
... A matriz de erros, bem como as métricas de exatidão global e o índice Kappa, vem sendo utilizada para determinar a acurácia de classificações digitais mediante a utilização de imagens de satélites (Jupp, 1989;Congalton & Green, 1999;De Wit & Clevers, 2004;Serra & Pons, 2008;Peña-Barragán et al., 2011). Essa matriz permite avaliar o desempenho da classificação realizada para uma classe individual, particularmente quando um pequeno número de classes de uso do solo é de interesse, como, por exemplo, na estimativa de área de uma cultura agrícola (Ceballos-Silva & López-Blanco, 2003). ...
Article
Full-text available
O objetivo deste trabalho foi estimar a área plantada com soja por meio da normalização da matriz de erros gerada a partir da classificação supervisionada de imagens TM/Landsat‑5. Foram avaliados oito municípios no Estado do Paraná, com dados referentes à safra de 2003/2004. As classificações foram realizadas por meio dos métodos paralelepípedo e máxima verossimilhança, dando origem à “máscara de soja”. Os valores do índice Kappa dos oito municípios ficaram acima de 0,6. As estimativas de área de soja, corrigidas por matriz de erros, apresentaram alta correlação com as estimativas oficiais do estado e com as estimativas geradas a partir de um método alternativo denominado “expansão direta”. A estimativa de área de soja por meio da normalização da matriz de erros apresenta menor custo e pode subsidiar métodos convencionais na estimativa menos subjetiva de safras.
... There are also studies that combine MODIS data and moderate spatial resolution data, such as Landsat and the Indian Remote Sensing Advanced Wide Field Sensor (AWIFS), to discriminate crop types USDA-NASS, 2013). Other studies used higher spatial resolution imagery, such as Landsat and ASTER data, to differentiate crop types for a less extensive area (Peña- Barragán et al., 2011;Serra and Pons, 2008;Turker and Arikan, 2005). Image selection for crop type mapping largely depends on the extensiveness of the study area, image availability, the cost, and the level of diversity in crop types and management. ...
Article
Site-specific information of crop types is required for many agro-environmental assessments. The study investigated the potential of support vector machines (SVMs) in discriminating various crop types in a complex cropping system in the Phoenix Active Management Area. We applied SVMs to Landsat time-series Normalized Difference Vegetation Index (NDVI) data using training datasets selected by two different approaches: stratified random approach and intelligent selection approach using local knowledge. The SVM models effectively classified nine major crop types with overall accuracies of >86% for both training datasets. Our results showed that the intelligent selection approach was able to reduce the training set size and achieved higher overall classification accuracy than the stratified random approach. The intelligent selection approach is particularly useful when the availability of reference data is limited and unbalanced among different classes. The study demonstrated the potential of utilizing multi-temporal Landsat imagery to systematically monitor crop types and cropping patterns over time in arid and semi-arid regions.
... For decades, many investigations have addressed the topic of crop discrimination via remote sensing through the use of various sources of satellite imagery, such as Landsat, QuickBird or Terra Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) [3][4][5]. In this topic, several authors have observed that information concerning variations in crop calendar, crop patterns, crop management techniques and parcel sizes shall be incorporated to the classifier algorithms for a successful result [6,7]. These variations can be derived from the textural, contextual or, in some cases, morphological features of the images [3][4][5]. ...
Article
Full-text available
The strategic management of agricultural lands involves crop field monitoring each year. Crop discrimination via remote sensing is a complex task, especially if different crops have a similar spectral response and cropping pattern. In such cases, crop identification could be improved by combining object-based image analysis and advanced machine learning methods. In this investigation, we evaluated the C4.5 decision tree, logistic regression (LR), support vector machine (SVM) and multilayer perceptron (MLP) neural network methods, both as single classifiers and combined in a hierarchical classification, for the mapping of nine major summer crops (both woody and herbaceous) from ASTER satellite images captured in two different dates. Each method was built with different combinations of spectral and textural features obtained after the segmentation of the remote images in an object-based framework. As single classifiers, MLP and SVM obtained maximum overall accuracy of 88%, slightly higher than LR (86%) and notably higher than C4.5 (79%). The SVM+SVM classifier (best method) improved these results to 89%. In most cases, the hierarchical classifiers considerably increased the accuracy of the most poorly classified class (minimum sensitivity). The SVM+SVM method offered a significant improvement in classification accuracy for all of the studied crops compared to the conventional decision tree classifier, ranging between 4% for safflower and 29% for corn, which suggests the application of object-based image analysis and advanced machine learning methods in complex crop classification tasks.
... A matriz de erros, bem como as métricas de exatidão global e o índice Kappa, vem sendo utilizada para determinar a acurácia de classificações digitais mediante a utilização de imagens de satélites (Jupp, 1989;Congalton & Green, 1999;De Wit & Clevers, 2004;Serra & Pons, 2008;Peña-Barragán et al., 2011). Essa matriz permite avaliar o desempenho da classificação realizada para uma classe individual, particularmente quando um pequeno número de classes de uso do solo é de interesse, como, por exemplo, na estimativa de área de uma cultura agrícola (Ceballos-Silva & López-Blanco, 2003). ...
Article
The objective of this work was to estimate soybean crop area by the normalization of the error matrix generated from the supervised classification of TM/Landsat‑5 images. Eight municipalities of the state of Paraná, Brazil, were evaluated using data from the 2003/2004 crop season. Classifications were carried out using the parallelepiped and maximum likelihood methods, resulting in a "soybean mask". Kappa index values for the eight municipalities were above 0.6. Estimated soybean areas, corrected by the error matrix, were highly correlated with official estimates of the state and with estimates generated from an alternative method called "direct expansion". Soybean crop area estimation by the normalization of the error matrix is less costly and can aid conventional methods in estimating harvests in a less subjective manner.
Preprint
Full-text available
Extreme rainfall and frequent flooding significantly impact hydro-geomorphic processes that shape the Earth's surface. This often induces increased sedimentation flow, which causes muck flow in the floodplains. It leads to changes in soil erosion/ deposition and flood extent. DEM of difference (DoD) and Elevation Deviation Index (EDI) analyses are widely used to study topographic changes in the earth’s surfaces due to soil erosion/deposition. However, vertical inaccuracies in the original DEMs are directly associated with the DoD, leading to errors in analyzing terrain changes. This paper presents a technique that minimizes the error by normalization, which allows a standardized comparison between the DEMs, making it easier to identify patterns and outliers. This is applied to the past 22 years of the data of the River Shilabati in West Bengal, India. The results of the study reveal a strong positive correlation between soil erosion and flood extent. As the region is frequently experiencing submersion by floodwaters with a simultaneous increase in the trend of extreme events, these findings have profound practical implications for flood management strategies.
Article
Jute crop holds significant economic importance contributing to environment and ecological balance. In the present study, optical and microwave satellite data was utilized to estimate jute acreage employing three distinct approaches, viz., logical criteria, machine learning and hybrid approach. Harmonized Sentinel-2 optical data products were utilized to generate monthly composite Enhanced Vegetation Index (EVI), whereas Ground Range Detected Sentinel-1 VH polarization Synthetic Aperture Radar data were used for generating monthly VH composite using median filter. Support Vector Machine (SVM), a machine learning algorithm, was adopted for this study. The EVI and VH multi-temporal profiles over the ground truth points were analysed and the criteria for identifying jute crop pixels were developed for different sowing period, i.e., March and April, during 2020 to 2023. The pre-processing of datasets and SVM was executed on the Google Earth Engine platform. The assessment of the three approaches highlighted variations in their predictive capabilities. Instead of its high classification accuracy (* 0.80), the logical approach has its limitations in terms of complex sets of criteria and adoptability over larger extent. On the contrary, SVM classifier could develop self-learning criteria, but overestimated the jute area with lower classification accuracy (Kappa * 0.45). The present study proposed a novel approach by coupling machine learning with simple logical criteria to address the limitations, while reserving its advantages. The analysis revealed that the capability of hybrid approaches in estimating the jute acreage was comparable with logical approach, with reduced operational complexities and wider adoptability over larger extent.
Chapter
Bangladesh is a global south hotspot due to climate change and sea level rise concerns. It is a highly disaster-prone country in the world with active deltaic shorelines. The shorelines are quickly changing to coastal accretion and erosion. Erosion is one of the water hazards to landmass sinking, and accretion related to land level rises due to sediment load deposition on the Bay of Bengal continental shelf. Therefore, this study aimed to explore shoreline status with change assessment for the three study years 1991, 2006, and 2021 using satellite remote sensing and geographical information system (GIS) approaches. Landsat 5, 7 (enhanced thematic mapper plus, ETM+), and 8 (operational land imager, OLI) satellite imageries were employed for onshore tasseled cap transformation (TCT) and land and sea classification calculations to create shore boundaries, baseline assessment, land accretion, erosion, point distance, and near-feature analysis. We converted 16,550 baseline vertices to points as the study ground reference points (GRPs) and validated those points using the country data sheet collected from the Survey of Bangladesh (SoB). We observed that the delta’s shorelines were changed, and the overall lands were accredited for the land-increasing characteristics analysis. The total accredited lands in the coastal areas observed during the time periods from 1991 to 2006 were 825.15 km2, from 2006 to 2021 were 756.69 km2, and from 1991 to 2021 were 1223.94 km2 for the 30-year period. Similarly, coastal erosion assessment analysis indicated that the results gained for the period 1991 to 2006 and 2006 to 2021 were 475.87 km2 and 682.75 km2, respectively. Therefore, the total coastal erosion was 800.72 km2 from 1991 to 2021. Neat accretion was 73.94 km2 for the 30-year period from 1991 to 2021. This research indicates the changes in shorelines, referring to the evidence for the delta’s active formation through accretion and erosion processes of “climate change” and “sea level rise.” This research projects the erosion process and threatens land use changes toward agriculture and settlements in the coastal regions of Bangladesh. The agricultural land use class was ranked one among other major land use and land cover (LULC) classes. In the study area, 34.03%, 34.47%, and 33.52% were occupied by agricultural land use for the study years 1991, 2006, and 2021, respectively, within the studied area of interest (AoI). Among all LULC patterns in the study, the agricultural lands increased in 2006 from the year 1991. On the other hand, the cultivated agricultural land had decreased in the year 2021 from the study year 2006. Hence, the agricultural lands were under great threat of loss of lands due to the coastal erosional processes in the Bangladesh delta. Therefore, it is urgent to develop new croplands area in the accredited new mud, soil, and intertidal (MSI) land-covered area for mitigating the loss of agricultural lands in onshore coastal areas of Bangladesh to strengthen the food security of the coastal region in Bangladesh.
Article
Full-text available
This contribution assesses a new term that is proposed to be established within Land Change Science: Spatio-TEmporal Patterns of Land (‘STEPLand’). It refers to a specific workflow for analyzing land-use/land cover (LUC) patterns, identifying and modeling driving forces of LUC changes, assessing socio-environmental consequences, and contributing to defining future scenarios of land transformations. In this article, we define this framework based on a comprehensive meta-analysis of 250 selected articles published in international scientific journals from 2000 to 2019. The empirical results demonstrate that STEPLand is a consolidated protocol applied globally, and the large diversity of journals, disciplines, and countries involved shows that it is becoming ubiquitous. In this paper, the main characteristics of STEPLand are provided and discussed, demonstrating that the operational procedure can facilitate the interaction among researchers from different fields, and communication between researchers and policy makers.
Thesis
High-dimensional hyperspectral (HS) remote sensing images are highly resourceful compared to multispectral (MS) data for different application but handling of such large volume data is a very challenging task, which should be addressed with the use of feature selection (FS) or feature extraction (FE)-based dimensionality reduction techniques. The research area considered in this thesis contributes towards developing new computationally efficient approaches based on the advanced machine learning and deep learning techniques to achieve better performance for land cover classification and chlorophyll content prediction. Prior to start the HS data application, MS data are also analyzed in this thesis for crop classification. First work considers surface reflectances and derived normalized difference indices (NDIs) of multi-temporal MS images (i.e. Landsat-8) combinedly for crop classification. FS and FE-based different dimensionality reduction techniques are employed on the multi-temporal datasets to detect the most favorable features, which are later utilized in the support vector machine (SVM) classifier to classify the crop types. In the second work, partial informational correlation (PIC)-based HS band selection approach is proposed as a FS-based dimensionality reduction technique for classification of land cover types. In this proposed approach, HS narrow-bands are divided into different spectral groups or segments using normalized mutual information (NMI) and then PIC is employed to each spectral group for optimal band selection. Further, these optimal spectral bands are used in the SVM and random forest (RF) classifiers for classification of land cover types. The third work proposes a computationally efficient FE-based dimensionality reduction approach, where NMI-based segmented stacked auto-encoder (S-SAE) are utilized as spectral features and consecutively used for creation of spatial features (i.e. extended morphological profiles (EMPs)), and later both spectral and spatial features are used in the classifier models (i.e. SVM and RF) for land cover classification. Fourth work addresses the issue of limited availability of space-borne HS data and proposes a deep learning-based regression algorithm (i.e. convolutional neural network regression (CNNR)) to transform MS data (i.e. Landsat 7/8) into quasi-HS (i.e. quasi-Hyperion) data. CNNR model introduces the advantages of nonlinear modelling and assimilation of spatial information in the regression-based modelling. The proposed CNNR model-based quasi-HS data quality is evaluated in terms of statistical metrics and crop classification application. The final work develops a model for estimation of canopy averaged chlorophyll content (CACC) of pear trees using the convolutional auto-encoder (CAE) features of HS data. This study also demonstrated the inspection of anomaly among the trees by employing multi-dimensional scaling (MDS) on the CAE features and detected the outlier trees, prior to fit nonlinear regression models (i.e. Gaussian process regression (GPR) and support vector regression (SVR)). The algorithms proposed in different chapters of the thesis provide better performances compared to the relevant existing techniques. The proposed research works (i.e. developed techniques and end products) have the potential applications in hydrological modelling, irrigation water management, crop yield forecasting etc. http://etd.iisc.ac.in/handle/2005/4537
Article
Full-text available
Recent advancements in the remotely sensed data products and machine learning algorithms are utilized effectively for classification of crops over a considerable large-area. This article proposes the use of feature extraction techniques to be employed on the multi-temporal Landsat-8 OLI sensor’s surface reflectances and derived Normalized Difference Indices datasets to classify different crop types. Numerous dimension reduction techniques, viz. feature selection (Random Forest and PIC measure based), linear (Principal Component Analysis (PCA) and Independent Component Analysis) and nonlinear feature extraction (kernel PCA and Autoencoder), are evaluated to detect most favourable features which should be apt for classification of crops. Subsequently, the detected features are used in a promising nonparametric classifier, support vector machine, for crop classification. It has been found that all the evaluated feature extraction techniques, employed on the multi-temporal datasets, result in better performance compared to feature selection based approaches. PCA, being a simple and efficient feature extraction algorithm, is well-suited in this classification study and extracted features can classify the crops with an average overall accuracy of 94.32%. Most of the crop types achieves user and producer accuracy of more than 90%. Multi-temporal images prove to be more advantageous compared to the single-date imagery for crop identification.
Article
Spectral index time series can provide valuable phenological information into the classification process for the precise crop mapping, in order to reduce misclassification rates associated with low interclass and high intraclass spectral variability. Stochastic hidden Markov models (HMMs) are efficient yet computationally demanding classification approach which can simulate crop dynamics, exploiting the spectral information of their phenological states and the relations between these states. This letter aims to present a methodology which achieves accurate classification results while maintaining a low computational cost. A classification framework based on HMMs was developed, and different spectral indices were generated from the time series of Landsat ETM+ and RapidEye imagery, for modeling crop vegetation dynamics over a Mediterranean rural area, with high spatiotemporal crop heterogeneity. To further improve the HMMs indices classification, separability analysis and two different decision fusion strategies were tested. The assessment of the classification accuracy, along with an evaluation of the computational cost, indicated that the green-red vegetation index produced the most favorable results among the individual spectral indices. Although the decision fusion based on an integration of a reliability factor increased the overall accuracy by 3.1%, this came at the cost of computational time, compared to the separability analysis model which required less processing time.
Article
Full-text available
An important metric to monitor for optimizing water use in agricultural areas is the amount of cropland left fallowed, or unplanted. Fallowed croplands are difficult to model because they have many expressions; for example, they can be managed and remain free of vegetation or be abandoned and become weedy if the climate for that season permits. We used 250 m, 8-day composite Moderate Resolution Imaging Spectroradiometer normalized difference vegetation index data to develop an algorithm that can routinely map cropland status (planted or fallowed) with over 75% user’s and producer’s accuracies. The Fallow-land Algorithm based on Neighborhood and Temporal Anomalies (FANTA) compares the current greenness of a cultivated pixel to its historical greenness and to the greenness of all cultivated pixels within a defined spatial neighborhood, and is therefore transportable across space and through time. This article introduces FANTA and applies it to California from 2001 to 2015 as a case study for use in data-poor places and for use in historical modeling. Timely and accurate knowledge of the extent of fallowing can provide decision makers with insights and knowledge to mitigate the impacts of drought and provide a scientific basis for effective management response. This study is part of the WaterSMART (Sustain and Manage America’s Resources for Tomorrow) project, an interdisciplinary and collaborative research effort focused on improving water conservation and optimizing water use. Download full article @: http://dx.doi.org/10.1080/15481603.2017.1290913
Chapter
The benefits of remote sensing for observing crops have been widely shown through the success that it has had in top producing countries. It is omnipresent today in national and global systems for predicting large industrial crop harvests and in precision farming services. However, such systems are still not common enough to describe and quantify the traditional production of developing countries. The 2009 report on world agriculture by the International Assessment of Agricultural Science and Technology for Development demonstrates the central role that smallholder farmers play in feeding the global population. Census counts indicate that there are nearly 500 million small farms around the world, and that in Africa 90% of agricultural production comes from small family farms. Trend analysis further suggests that small farms are expected to continue dominating agricultural landscapes in developing countries, especially in Asia and Africa, for at least the next two or three decades. Moreover, because of new land subdivisions and the cultivation of new areas, the number of these farms continues to increase in many countries.
Article
In this research, we evaluated three classifiers for crop identification in multi-temporal images, in northwest of Iran. These classifiers are maximum likelihood, support vector machine and random forest (the two latter are machine learning classifiers). Our data consist of a two-date SPOT5 image for early spring and mid-summer. There are 7 crops cultivated in this region, including corn, rice, rain-fed wheat, irrigated wheat, fodder, summer crops and fallow. We test three classifier with a two-date image and once with each of these one-date images. Our results indicated that SVM classifier performed best in comparison with two other classifiers in all images (overall accuracy in two-date image with SVM is 80.9%, with RF 80.11% and with ML 77.05%). However, RF classifier performance in two-date image is comparable with SVM. In all classifications, the early-spring image has lowest accuracy and two-date image has highest accuracy.
Article
Full-text available
This study proposes an empirical methodology for modelling and mapping the air temperature (mean maximum, mean and mean minimum) and total precipitation, all of which are monthly and annual, using geographical information systems (GIS) techniques. The method can be seen as an alternative to classical interpolation techniques when spatial information is available. The geographical area used to develop and apply this model is Catalonia (32000 km 2 , northeast Spain). We have developed a multiple regression analysis between these meteorological variables as the dependent ones, and some geographical variables (altitude (ALT), latitude (LAT), continentality (CON), solar radiation (RAD) and a cloudiness factor (CLO)) as the independent ones. Data for the dependent variables were obtained from meteorological stations, and data for the independent variables were elaborated from a 180 m resolution digital elevation model (DEM). Multiple regression coefficients (b n) were used to build final maps, using digital layers for each independent variable, and applying basic GIS techniques. The results are very satisfactory in the case of mean air temperature and mean minimum air temperature, with coefficients of determination (R 2) between 0.79 and 0.97, depending on the month; in the case of mean maximum air temperature, R 2 ranges between 0.70 and 0.89, while in the case of precipitation, it ranges between 0.60 and 0.91.
Article
Full-text available
This paper summarizes the consequences in the area classified and in the thematic accuracy of being more or less conservative in a hybrid classifier. The most important parameter of that classification consists in fidelity (the introduction of the threshold proportion at which to accept a spectral class as being a part of a thematic category). Two options have been tested: the first less conservative, the second more conservative. These fidelities have been applied to ten Mediterranean crops and tested using error matrices. Thematic accuracies were quantified following the classical approach (number of pixels correctly classified), a polygon approach (number of polygons correctly classified) and, finally, area approach (area correctly classified). Results showed that the most restrictive fidelity produces less area classified but with more thematic accuracy when unclassified pixels are not included in the quantification of the accuracy. This fact occurred in all the options (pixel, polygon and area) although did not affect all the crops equally.
Article
Full-text available
The need for multi-temporal data analysis for delineation of wheat crop has been demonstrated first. It is found that Maximum Likelihood Classification (MLC) with the composite data of multi-temporal images is limited by the problem of large null set containing crop pixels. Therefore, for effective classification of multi-temporal images, two approaches are evaluated: (1) MLC with different strategies--sequential MLC (s_MLC), MLC with Principal Components (pca_MLC) and iterative MLC (i_MLC); and (2) Artificial Neural Networks (ANN) with back-propagation method. These classifiers were applied on multi-temporal Indian Remote Sensing satellite (IRS)-1B images to classify wheat crop in two areas of India, one with dominant wheat and the other with less dominant wheat cultivation. Among the three strategies of MLC, i_MLC has resulted in relatively better classification of wheat. However, the result of ANN classification is superior to that of i_MLC with respect to the correctness of labelling of wheat pixels. The performance of ANN is proved to be better, in both the situations of dominant wheat and less dominant wheat cultivation.
Article
Full-text available
This study focuses on the geometrical deformations introduced by relief in images captured by the TM sensor of Landsat satellites and by the HRV sensor of SPOT satellites. Different correction alternatives are presented in order to incorporate altitude data into correction procedures based on first-degree polynomial models. Column and row determinations from the corresponding map coordinates are carried out independently. Three different models for columns and two for rows are proposed. The results have been contrasted with those obtained using classic first- and second-degree polynomial calculations, and with those obtained using an orbital model (for SPOT images). The models presented are easy to implement and provide a level of precision similar to that of the orbital model used, while they are much more efficient in calculation time. In view of the results, the model which integrates altimetric data into a single first-degree polynomial seems of particular interest.
Article
Full-text available
The objective of this study was to explore the use of multi-temporal Landsat TM data from the same growing season for the classification of land cover types in the south-western portion of the Argentine Pampas. Investigations were made on how many dates are necessary to obtain an accurate classification and, given a fixed number of dates, which is the particular combination of dates that yield the best results. Additionally, the effect of using the NDVI instead of all the bands available on the classification accuracy and the use of a moving window filter over the classified image were tested. Scenes acquired in spring, early summer, late summer and early fall of the 1996-1997 growing season were used. Land cover information for the same period was collected from farms and ranches and this information was included in a GIS. Supervised classifications were performed using all the 15 possible ways to combine the four dates. At least two scenes are needed for a satisfactory classification. These scenes must embrace the shift between winter and summer crops (i.e. one spring and one summer image). Using the NDVI instead of Landsat TM bands 3, 4 and 5 increased the biological interpretability of the results but caused a decrease in accuracy.
Article
Full-text available
Change detection from remote sensing data is often done by simple overlay of classified maps. However, such analyses can contain a significant proportion of boundary errors, especially when combining data from different sensors. This paper presents a protocol that allows reliable post-classification comparisons by taking into account classification accuracies, landscape fragmentation, planimetric accuracies, pixel sizes and grid origins. The proposed protocol has been applied, with little extra effort, in a fragmented agricultural Mediterranean zone using MSS (1970s) and TM (1990s) images. Applying the protocol, change detection had an accuracy of 85.1%, while for a direct overlay it was only 43.9% accurate. The drawback of this method is that it reduces the useful area of comparison. As the accuracy of individual classifications is critical, the paper also describes and tests a hybrid classifier that combines an unsupervised classification approach with training areas. This approach has proved more successful than maximum likelihood classifiers.
Article
Full-text available
A new tasseled cap transformation based on Landsat 7 at-satellite reflectance was developed. This transformation is most appropriate for regional applications where atmospheric correction is not feasible. The brightness, greenness and wetness of the derived transformation collectively explained over 97% of the spectral variance of the individual scenes used in this study. Keyword: tasseled cap transformation, at-satellite reflectance, and Landsat 7. Running headline: At-satellite reflectance based tasseled cap transformation.
Article
Full-text available
The use of thematic mapper data to study crop growth parameters has primarily been conducted with hand-held and truck-mounted radiometers. In addition, most studies have not used mid-infrared bands to develop vegetation indices. This study used Landsat thematic mapper data and crop parameter data from corn and soybean fields in northern Ohio (1) to develop plant-water sensitive vegetation indices that included the mid-infrared, near-infrared, and red bands; and (2) to compare the ability of different vegetation indices to estimate crop yield, leaf area index, wet biomass, dry biomass, and plant height. Indices which included the mid-infrared thematic mapper bands 5 and 7 performed equal to or better than the widely used near-infrared band 4 and red band 3. The best relationships between vegetation indices and crop parameters had Spearman non-parametric correlation coefficients of 0.80 to 0.89 for all parameters except crop yield which was not highly correlated with the vegetation indices.
Article
Full-text available
The earlier generation of irrigation performanceindicators was based on canal flow data. Commonly,they quantify performance in a command area downstreamof a discharge measurement device. Remote sensingdeterminants, such as actual evapo-transpiration, soilwater content and crop growth reflect the overallwater utilization at a range of scales, up to fieldlevel. Crop evapo-transpiration includes wateroriginating from irrigation supply, water fromprecipitation, groundwater and water withdrawn fromthe unsaturated zone. Hence, this is a refinement inspatial scale as compared to the classicallycollected flow measurements, and describes moreoverdepletion from all water resources. If thesepossibilities are well implemented, we expect that anew generation of irrigation performance indicatorscan be quantified in a cost-effective manner.Especially, because satellite measurements pave a wayto standardize data collection between differentirrigation schemes and among different countries atcosts which are currently decreasing. These challengescan only turn into a success if irrigation managersare involved in pilot projects and demonstrationstudies exploring satellite data.
Article
Full-text available
Digital analysis of remotely sensed data has become an important component of many earth-science studies. These data are often processed through a set of preprocessing or “clean-up” routines that includes a correction for atmospheric scattering, often called haze. Various methods to correct or remove the additive haze component have been developed, including the widely used dark-object subtraction technique. A problem with most of these methods is that the haze values for each spectral band are selected independently. This can create problems because atmospheric scattering is highly wavelength-dependent in the visible part of the electromagnetic spectrum and the scattering values are correlated with each other. Therefore, multispectral data such as from the Landsat Thematic Mapper and Multispectral Scanner must be corrected with haze values that are spectral band dependent. An improved dark-object subtraction technique is demonstrated that allows the user to select a relative atmospheric scattering model to predict the haze values for all the spectral bands from a selected starting band haze value. The improved method normalizes the predicted haze values for the different gain and offset parameters used by the imaging system. Examples of haze value differences between the old and improved methods for Thematic Mapper Bands 1, 2, 3, 4, 5, and 7 are 40.0, 13.0, 12.0, 8.0, 5.0, and 2.0 vs. 40.0, 13.2, 8.9, 4.9, 16.7, and 3.3, respectively, using a relative scattering model of a clear atmosphere. In one Landsat multispectral scanner image the haze value differences for Bands 4, 5, 6, and 7 were 30.0, 50.0, 50.0, and 40.0 for the old method vs. 30.0, 34.4, 43.6, and 6.4 for the new method using a relative scattering model of a hazy atmosphere.
Article
Full-text available
The normalized difference vegetation index (NDVI) has been widely used for remote sensing of vegetation for many years. This index uses radiances or reflectances from a red channel around 0.66 μm and a near-IR channel around 0.86 μm. The red channel is located in the strong chlorophyll absorption region, while the near-IR channel is located in the high reflectance plateau of vegetation canopies. The two channels sense very different depths through vegetation canopies. In this article, another index, namely, the normalized difference water index (NDWI), is proposed for remote sensing of vegetation liquid water from space. NDWI is defined as , where ϱ represents the radiance in reflectance units. Both the 0.86-μm and the 1.24-μm channels are located in the high reflectance plateau of vegetation canopies. They sense similar depths through vegetation canopies. Absorption by vegetation liquid water near 0.86 μm is negligible. Weak liquid absorption at 1.24 μm is present. Canopy scattering enhances the water absorption. As a result, NDWI is sensitive to changes in liquid water content of vegetation canopies. Atmospheric aerosol scattering effects in the 0.86–1.24 μm region are weak. NDWI is less sensitive to atmospheric effects than NDVI. NDWI does not remove completely the background soil reflectance effects, similar to NDVI. Because the information about vegetation canopies contained in the 1.24-μm channel is very different from that contained in the red channel near 0.66 μm, NDWI should be considered as an independent vegetation index. It is complementary to, not a substitute for NDVI. Laboratory-measured reflectance spectra of stacked green leaves, and spectral imaging data acquired with Airborne Visible Infrared Imaging Spectrometer (AVIRIS) over Jasper Ridge in California and the High Plains in northern Colorado, are used to demonstrate the usefulness of NDWI. Comparisons between NDWI and NDVI images are also given.
Article
A methodology for supervised classification to identify irrigated crops with Landsat TM imagery in a semiarid zone (La Mancha, Spain) is presented. The discrimination procedure is based on the different crop spectral responses through time according to their phenological evolution. Our multitemporal supervised classification includes maximum-likelihood algorithms, decision-tree criteria, and context classifiers. We have applied the procedure to two sets of scenes obtained for the growing seasons of 1996 and 1997, respectively. The resulting classification accuracy was 93.1 percent for 1996 and 90.21 percent for 1997. We have estimated the areas occupied by each crop class by means of intersecting the TM-derived land-use raster map and the digital rural cadastre vector map in a geographic information system. We have assessed the accuracy of the crop area estimation from the classified image by comparing these areas with those calculated from the digital rural cadastre. A median filter applied to the final classification improves the agreement of the estimated crop areas with the cadastre data. Additional post-classification methods to correct crop areas did not bring any significant further improvements. Therefore, we conclude that the context classifier is a useful and sufficient tool to improve surface quantification.
Article
Multispectral scanner system data simulating the TM of Landsat-4 were analysed for an area near Gedney Hill, Lincolnshire, UK. The data were found to have a three-dimensional statistical structure similar to that for the Landsat-4 TM of parts of the US. Divergence analysis indicates that the optimal choice of bands for cover discrimination should include one band from the visible, one from the near-IR and one from the middle- or far-IR. It was further shown, primarily from consideration of principal-component images, that significant discriminatory power may be lost if all bands are not used. Comparisons with Landsat-4 TM principal-component images are made. The role of noise factors in obscuring information, especially from highly correlated bands, is shown to be of considerable importance. -Author
Article
Multi-temporal Landsat 5 Thematic Mapper (TM) imagery was evaluated for the identification and monitoring of potential jurisdictional wetlands located in the states of Maryland and Delaware. A wetland map prepared from single-date TM imagery was compared to a hybrid map developed using two dates of imagery. The basic approach was to identify land-cover vegetation types using spring leaf-on imagery, and identify the location and extent of the seasonally saturated soil conditions and areas exhibiting wetland hydrology using spring leaf-off imagery. The accuracy of the wetland maps produced from both single- and multiple-date TM imagery were assessed using reference data derived from aerial photographic interpretations and field observation data. Subsequent to the merging of wetland forest and shrub categories, the overall accuracy of the wetland map produced from two dates of imagery was 88 percent compared to the 69 percent result from single-date imagery. A Kappa Test Z statistic of 5.8 indicated a significant increase in accuracy was achieved using multiple-date TM images. Wetland maps developed from multi-temporal Landsat TM imagery may potentially provide a valuable tool to supplement existing National Wetland Inventory maps for identifying the location and extent of wetlands in northern temperate regions of the United States.
Article
The application of multispectral scanner (MSS) data to vegetation and soils studies can be facilitated by use of data transformations which reduce the number of channels to be considered, provide a more direct association between signal response and physical processes on the ground, and highlight the particular types of information of greatest interest to the user. One such transformation, the TM (Thematic Mapper) Tasseled Cap transformation, is described. Previously reported results are summarized and the results of new analyses pertaining to vegetation, soils, and external effects information contained in the TM Tasseled Cap feature space are presented.
Article
The Landsat-7 ETM+ sensor offers several enhancements over the Landsat-4,5 Thematic Mapper (TM) sensor, including increased spectral information content, improved geodetic accuracy, reduced noise, reliable calibration, the addition of a panchromatic band, and improved spatial resolution of the thermal band. In this paper, we present some initial comparisons between Landsat-5 TM and Landsat-7 ETM+ imagery in order to quantify these improvements. We find that the ETM+ continues the record of TM observations, and, in many respects, substantially improves upon the earlier sensor. Specific improvements include lower spatial noise levels, improved information content, and geodetic accuracy of systematically corrected products to 50–100 m. These improvements are likely to have significant benefits for land-cover mapping and change detection applications.
Article
Multispectral scanner system data simulating the thematic mapper (TM) of LANDSAT-4 were analysed for an area near Gedney Hill, Lincolnshire, U.K. The data were found to have a three-dimensional statistical structure similar to that for the LANDSAT-4 TM of parts of the United States. Divergence analysis indicates that the optimal choice of bands for cover discrimination should include one band from the visible, one from the near-IR (infrared) and one from the middle- or far-IR. It was further shown, primarily from consideration of principal-component images, that significant discriminatory power may be lost if all bands are not used. Comparisons with LANDSAT-4 TM principal-component images are made. The role of noise factors in obscuring information, especially from highly correlated bands, is shown to be of considerable importance.
Article
The imaging frequency and synoptic coverage of the National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) make possible for the first time a phenological approach to vegetation cover classification in which classes are defined in terms of the timing, the duration and the intensity of photosynthetic activity. This approach, which exploits the strong, approximately linear relationship between the amount of solar irradiance absorbed by plant pigments and shortwave vegetation indices calculated from red and near-infrared reflectances, involves a supervised binary decision tree classification of phytophenological variables derived from multidate normalized difference vegetation index (NDVI) imagery. A global phytophenological classification derived from NOAA global vegetation index imagery is presented and discussed. Although interpretation of the various classes is limited considerably by the quality of global vegetation index imagery, the data show clearly the marked temporal asymmetry of terrestrial photosynthetic activity.
Article
Population growth, urban expansion, land degradation, civil strife and war may place plant natural resources for food and agriculture at risk. Crop and yield monitoring is basic information necessary for wise management of these resources. Satellite remote sensing techniques have proven to be cost-effective in widespread agricultural lands in Africa, America, Europe and Australia. However, they have had limited success in Mediterranean regions that are characterized by a high rate of spatio-temporal ecological heterogeneity and high fragmentation of farming lands. An integrative knowledge-based approach is needed for this purpose, which combines imagery and geographical data within the framework of an intelligent recognition system. This paper describes the development of such a crop recognition methodology and its application to an area that comprises approximately 40% of the cropland in Israel. This area contains eight crop types that represent 70% of Israeli agricultural production. Multi-date Landsat TM images representing seasonal vegetation cover variations were converted to normalized difference vegetation index (NDVI) layers. Field boundaries were delineated by merging Landsat data with SPOT-panchromatic images. Crop recognition was then achieved in two-phases, by clustering multi-temporal NDVI layers using unsupervised classification, and then applying ‘split-and-merge’ rules to these clusters. These rules were formalized through comprehensive learning of relationships between crop types, imagery properties (spectral and NDVI) and auxiliary data including agricultural knowledge, precipitation and soil types. Assessment of the recognition results using ground data from the Israeli Agriculture Ministry indicated an average recognition accuracy exceeding 85% which accounts for both omission and commission errors. The two-phase strategy implemented in this study is apparently successful for heterogeneous regions. This is due to the fact that it allows unsupervised classification to represent the high phenological variability (by utilizing 70 clusters). Utilization of the ‘split-and-merge’ rules derived from the entire data set of imagery and auxiliary data enabled the formalization of different interpretation contexts for each crop. This technique, which uses imagery information in both stages, is significantly different from exiting methods that are based only on auxiliary geographical and expert knowledge in the post-classification phase.
Article
Nine scenes of SPOT/HRV data obtained in eight different months in 1997 were evaluated for crop discrimination in the Saga Plains, Japan. All images were atmospherically corrected with the 6S code. Annual Normalized Difference Vegetation Index (NDVI) profiles were generated to characterize seasonal trends in six cropping systems (rice, rice-winter cereal, soybean, soybean-winter cereal, lotus, and rush). The dataset of this study showed the unique temporal change patterns of NDVI for each cropping system. Separability analyses determined optimal scene combinations for the highest accuracy in classifying the cropping systems. The scene combinations for the accurate classification of cropping systems were obtained from three separability measurements (Euclidean spectral distance, divergence, and Jeffries-Matsushita distance). Kappa statistics were applied to evaluate the classification accuracies. The four-scene combination that was derived from April, June, July and September classified the cropping systems almost as well as those combinations including more scenes. A colour composition technique applied to the three-scene combination that showed the highest separability also discriminated each cropping system. Based on these results, we can request observations during specific time intervals considering local crop calendars and environmental conditions.
Article
The agricultural land cover in a 263 km2 irrigation district was classified utilizing two Landsat 5 TM scenes. Manual and automatic selection of training areas for the classification of two single subscenes and a combined multitemporal subscene result in several differently classified images. The extent of each land cover class was first estimated by area frame sampling and further expansion of the ground data to the entire irrigation area. The regression of the sampled surface on the corresponding pixels in the classified images was used to improve the regression estimates of the areas of different land cover classes. To ascertain if there is any statistical difference between the relative efficiencies (RE) of the regression estimator using each one of the classifications, and being RE= 1/ (1 -r) a test of equality between correlation coefficients was applied. When the correlation coefficients were significantly different the most precise estimation was indicated by the highest RE. Manual multitemporal classifications provided the most precise results, with the exception of the single image spring classification of rice.
Article
A process for integrating remote sensing and spatial data analysis to accurately map and monitor agricultural crops and other land cover in the Lower Colorado River Basin is described. These maps were then used as input into a model that accounts for consumptive use throughout the basin. Water is an important and incredibly valuable resource in this area. International treaties and court decrees dictate water allocation to the states of Arizona, Nevada, and California, and to Mexico. Maps of the agricultural crops with a required overall accuracy of 93 percent for use in the water model were generated from Landsat Thematic Mapper data four times per year. An automated signature extraction process and data exploration techniques were developed to aid in achieving these required accuracies. All maps were subjected to quantitative accuracy assessment, and error matrices were produced to evaluate overall and per-class accuracies.
Article
Water availability is the key factor determining maize yields in NE Spain. Irrigation is needed to obtain economic yields but it is costly and water supply is sometimes insufficient. The aim of this research was to test a simple simulation model for evaluating different irrigation strategies, especially under water-limited conditions. The LINTUL model was adapted and parameterized using experimental data from the 1995 season. Most parameters were obtained from experiments, although some were taken from the literature. This model is based on the concept of light use efficiency, incorporates a soil water balance and simulates phenology, crop leaf area, biomass accumulation and yield. It was tested on independent data from the 1995 and 1996 seasons under different irrigation treatments. The model predicted the flowering date within ±5 days of the observed values. Leaf area index was predicted satisfactorily, except under extreme water-stress conditions, where it was overestimated. In general, soil moisture content and yield were accurately predicted. In the 1996 experiment measured yields ranged from 6.4 to 13.6 t ha−1 and simulated yields from 6.5 to 12.2 t ha−1. These results show that the LINTUL model can be used as a tool for exploring the consequences on maize yields of different irrigation strategies in NE Spain. Analysis of the model identified a process that strongly affects yield loss due to drought, but for which present understanding is still insufficient: the effects of drought on leaf senescence and canopy architecture.
Article
The Landsat-7 ETM+ sensor offers several enhancements over the Landsat-4,5 Thematic Mapper (TM) sensor, including increased spectral information content, improved geodetic accuracy, reduced noise, reliable calibration, the addition of a panchromatic band, and improved spatial resolution of the thermal band. In this paper, we present some initial comparisons between Landsat-5 TM and Landsat-7 ETM+ imagery in order to quantify these improvements. We find that the ETM+ continues the record of TM observations, and, in many respects, substantially improves upon the earlier sensor. Specific improvements include lower spatial noise levels, improved information content, and geodetic accuracy of systematically corrected products to 50–100 m. These improvements are likely to have significant benefits for land-cover mapping and change detection applications.
Article
A simplified model for radiometric corrections has been used to improve nonsupervised classification of vegetation cover in a hilly area near Barcelona, Spain. A digital elevation model and standard parameters for exoatmospheric solar irradiance, atmospheric optical depth, and sensor calibration are the only inputs required. Radiometric classes obtained by cluster classification of Landsat TM images from nonradiometrically corrected images include several classes related to terrain illumination, but not to vegetation or thematic cover differences. The use of radiometric correction allows identifying all radiometric classes obtained as vegetation or thematic classes with 83.3% global accuracy. Classes obtained include Pinus halepensis, Quercus ilex, and Quercus cerrioides forests, shrublands, grasslands, urban areas with vegetation, urban areas without vegetation, and denuded areas. Radiometric correction helps in estimating surfaces and spectral features of these classes. The results are discussed considering botanical composition, date (phenology), and vegetation dynamics.
Article
This popular text introduces students to widely used forms of remote sensing imagery and their applications in plant sciences, hydrology, earth sciences, and land use analysis. Providing comprehensive coverage of principal topics in the field, the book's 4 sections and 21 chapters are carefully designed as independent units that instructors can select from as needed for their courses. Relevant case studies and review questions that reinforce the concepts presented in each chapter make this book essential reading for students in remote sensing. Illustrations include 28 color plates and nearly 400 black-and-white images and figures.
Article
Information about vegetation water content (VWC) has widespread utility in agriculture, forestry, and hydrology. It is also useful in retrieving soil moisture from microwave remote sensing observations. Providing a VWC estimate allows us to control a degree of freedom in the soil moisture retrieval process. However, these must be available in a timely fashion in order to be of value to routine applications, especially soil moisture retrieval. As part of the Soil Moisture Experiments 2002 (SMEX02), the potential of using satellite spectral reflectance measurements to map and monitor VWC for corn and soybean canopies was evaluated. Landsat Thematic Mapper and Enhanced Thematic Mapper Plus data and ground-based VWC measurements were used to establish relationships based on remotely sensed indices. The two indices studied were the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI). The NDVI saturated during the study period while the NDWI continued to reflect changes in VWC. NDWI was found to be superior based upon a quantitative analysis of bias and standard error. The method developed was used to map daily VWC for the watershed over the 1-month experiment period. It was also extended to a larger regional domain. In order to develop more robust and operational methods, we need to look at how we can utilize the MODIS instruments on the Terra and Aqua platforms, which can provide daily temporal coverage.
Article
A knowledge-based classification method was designed to improve crop classification accuracy. Crop data of preceding years, stored in a geographical information system (GIS) were used as ancillary data. Knowledge about crop succession, determined from crop rotation schemes, was formalized by means of transition matrices. The spectral data, the data from the GIS and the knowledge represented in the transition matrix were used in a modified Bayesian classification algorithm. The developed classification was tested in an agricultural region in The Netherlands. Depending on the spectral class discrimination, the accuracy of the knowledge-based classification was 6 to 20 percent better compared with a maximum likelihood classification.
Article
A model, utilizing direct relationship between remotely sensed spectral data and the development stage of both corn and soybeans has been proposed and published previously (Badhwar and Henderson, 1981; and Henderson and Badhwar, 1984). This model was developed using data acquired by instruments mounted on trucks over field plots of corn and soybeans as well as satellite data from Landsat. In all cases, the data was analyzed in the spectral bands equivalent to the four bands of Landsat multispectral scanner (MSS). In this study the same model has been applied to corn and soybeans using Landsat-4 Thematic Mapper (TM) data combined with simulated TM data to provide a multitemporal data set in TM band intervals. All data (five total acquisitions) were acquired over a test site in Webster County, Iowa from June to October 1982. The use of TM data for determining development state is as accurate as with Landsat MSS and field plot data in MSS bands. The maximum deviation of 0.6 development stage for corn and 0.8 development stage for soybeans is well within the uncertainty with which a field can be estimated with procedures used by observers on the ground in 1982.
Article
The time trajectories of agricultural data points as seen in Landsat signal space form a pattern suggestive of a tasselled woolly cap. Most of the important crop phenomena can be described using this three dimensional construct: the distribution of signals from bare soil, the processes of green development, yellow development, and shadowing and harvesting. A linear preprocessing transformation which isolates green development, yellow development and soil brightness is used to reduce the dimension of the signal space. Specific measurable pattern elements of the tasselled cap are used to estimate and correct atmospheric haze and moisture effects.
Remote sensing and GIS for site-specific farming
  • J G Lyon
  • A Ward
  • B C Atherton
  • G Senay
  • T Krill
LYON, J.G., WARD, A., ATHERTON, B.C., SENAY, G. and KRILL, T., 2003, Remote sensing and GIS for site-specific farming. In GIS for water resources and watershed management, J.G. Lyon (Ed.) (Boca Raton: CRC Press).
Modelling moisture content of fine dead wildland fuels: input to the BEHAVE fire prediction system Intermountain Research Station, Research Paper INT-359 Weighting fidelity versus classified area in remote sensing classifications from a pixel and a polygon perspective
  • R Wilson
  • R A Morris
  • G A Sackett
  • G Pons
ROTHERMEL, R., WILSON, R.A., MORRIS, G.A. and SACKETT, S.S., 1986, Modelling moisture content of fine dead wildland fuels: input to the BEHAVE fire prediction system, USDA Forest Service, Intermountain Research Station, Research Paper INT-359, Ogden, Utah. SERRA, P., MORÉ, G. and PONS, X., 2006, Weighting fidelity versus classified area in remote sensing classifications from a pixel and a polygon perspective. In Proceedings of 7th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences, M. Caetano and M. Painho (Eds).
Calendario de siembra, floració y recolecció
  • Ministerio De Agricultura
  • Pesca Alimentacio
MINISTERIO DE AGRICULTURA, PESCA Y ALIMENTACIO N, 1982, Calendario de siembra, floració y recolecció (Madrid: Ministerio de Agricultura).
Land Cover Map of Catalonia Available online at: http://www.creaf.uab.es/usa/ productes An improved dark-object subtraction technique for atmospheric scattering correction of multispectral data. Remote Sensing of Environment
  • Center For Ecological Research And Forestry
  • Applications
CENTER FOR ECOLOGICAL RESEARCH AND FORESTRY APPLICATIONS (CREAF), 2006, Land Cover Map of Catalonia. Available online at: http://www.creaf.uab.es/usa/ productes.htm. CHAVEZ, P.S., 1988, An improved dark-object subtraction technique for atmospheric scattering correction of multispectral data. Remote Sensing of Environment, 24, pp. 459–479.
Geographic Information System and Remote Sensing Software, Centre de Recerca Ecològica i Aplicacions Forestals, CREAF, Bellaterra Available online at A simple radiometric correction model to improve automatic mapping of vegetation from multispectral satellite data. Remote Sensing of Environment
  • X Pons
  • X Pons
  • L L Solé-Sugran Es
PONS, X., 2002, MiraMon. Geographic Information System and Remote Sensing Software, Centre de Recerca Ecològica i Aplicacions Forestals, CREAF, Bellaterra. Available online at: http://www.creaf.uab.es/miramon. PONS, X. and SOLÉ -SUGRAN ES, L.L., 1994, A simple radiometric correction model to improve automatic mapping of vegetation from multispectral satellite data. Remote Sensing of Environment, 48, pp. 191–204.
Digital climatic atlas of Catalonia Available online at: http://magno.uab.es/atles-climatic/en_index Incorporation of relief in polynomial-based geometric corrections
  • M Ninyerola
  • X Pons
  • J M Htm
  • V Pala
  • X Pons
NINYEROLA, M., PONS, X. and ROURE, J.M., 2004, Digital climatic atlas of Catalonia. Available online at: http://magno.uab.es/atles-climatic/en_index.htm. PALA, V. and PONS, X., 1995, Incorporation of relief in polynomial-based geometric corrections. Photogrammetric Engineering and Remote Sensing, 61, pp. 935–944.
Digital climatic atlas of Catalonia
  • M Ninyerola
  • X Pons
  • J M Roure
NINYEROLA, M., PONS, X. and ROURE, J.M., 2004, Digital climatic atlas of Catalonia. Available online at: http://magno.uab.es/atles-climatic/en_index.htm.
Land Cover Map of Catalonia
  • Center
  • Applications
CENTER FOR ECOLOGICAL RESEARCH AND FORESTRY APPLICATIONS (CREAF), 2006, Land Cover Map of Catalonia. Available online at: http://www.creaf.uab.es/usa/ productes.htm.
Pesca y Alimentación (1982) Calendario de siembra, floración y recolección
  • Ministerio De Agricultura
Modelling moisture content of fine dead wildland fuels: input to the BEHAVE fire prediction system
  • R Rothermel
  • R A Wilson
  • G A Morris
  • S S Sackett
ROTHERMEL, R., WILSON, R.A., MORRIS, G.A. and SACKETT, S.S., 1986, Modelling moisture content of fine dead wildland fuels: input to the BEHAVE fire prediction system, USDA Forest Service, Intermountain Research Station, Research Paper INT-359, Ogden, Utah.
  • T M Kiefer
LILLESAND, T.M., KIEFER, R.W. and CHIPMAN, J.W., 2004, Remote Sensing and Image Interpretation (USA: Wiley & Sons).
Calendario de siembra, floración y recolección (Madrid: Ministerio de Agricultura)
  • Pesca Ministerio De Agricultura
  • Alimentació N
MINISTERIO DE AGRICULTURA, PESCA Y ALIMENTACIÓ N, 1982, Calendario de siembra, floración y recolección (Madrid: Ministerio de Agricultura).
  • Lyon J. G.
  • Rothermel R.