ArticlePDF Available

Hypertemporal image analysis for crop mapping and change detection

Authors:

Abstract

Many authors explored the use of multi-temporal images, recorded within a season or across years, for (i) ecosystem monitoring, (ii) land cover (crop) identification, and (iii) change detection (Copin et.al, 2004). Temporary trajectory analysis, drawing on time-profile-based data originating from a large number of observation dates, has mainly been done through threshold-based methods, compositing-algorithms, or Fourier series approximation. This paper presents findings of a multivariate change detection method that processes the full dimensionality (spectral and temporal) of 10-day composite (1998-onwards) 1-km resolution SPOT-Vegetation NDVI images. Using the ISODATA clustering algorithm of Erdas-Imagine software and all available NDVI image data layers, unsupervised classification runs were carried out. These produced minimum-and average-divergence statistical indicators that in turn were used to identify the optimum number of classes that best suited the data put to the unsupervised classification algorithm. The selected classified map is linked to a set of time-profile-based signatures (profiles) that form the map legend. Studies were carried out for (i) Portugal to identify the extend and nature of land cover units, (ii) the Limpopo valley, Mozambique to map gradients, (iii) the Limpopo valley, Mozambique, to monitor flooded areas, (iv) Garmsar, Iran, to detect spatial differences in water availability, (v) Nizamabad, India, to link NDVI profiles to land use classes and (vi) Andalucía, Spain to disaggregate reported agricultural crop statistics to 1x1km pixel crop maps. Results compose of statistical findings underpinning the method, maps showing the spatial-temporal characteristics of the findings, and the applicability of the method for the studied topics.
A preview of the PDF is not available
... Warldow and Egbert (2011) used MODIS images for their approach and found some problems of accuracy that can be easily solved through using Landsat images. De Biel et al (2006) met a similar problem due to the use of SPOT images. These as well can be solved through the use of Landsat images. ...
... De Biel et al (2006) had already worked on the same methodology but did it using SPOT 4 images. Using the ISODATA clustering algorithm, unsupervised classification was carried out. ...
... After a detailed and comprehensive literature review, we have found that using Landsat images in order to map crops, proves the most efficient in our case as it proved the most suitable for our relatively small country field and limited resources and budget (Qinghan et al (2008), Ulbricht et al (2013), Warldow and Egbert (2011), De Biel et al (2006) Awad and Darwish (2011)). It was also clear that using the multi-temporal (Yas et. ...
Thesis
Nowadays crop management works on using techniques and practices to enhance production and to optimize the use of natural resources. Most developing countries need reliable techniques which can help in increasing production to meet the demand for food and to reduce over importation. These techniques are based on efficient mapping, planning for productive crops, and detecting problems ahead. In this research, geospatial technologies including remote sensing, Geographic Information System (GIS) and Global Positioning System (GPS) are used. In addition, a new expert based classification model based on the Normalized Differences Vegetation Index (NDVI) and supervised and unsupervised image classification algorithms are used to create crop maps of Lebanon. Mapping are focused on major crops such as wheat, potato, vineyard, fruit trees, and other major crop types. These crops are selected because they are the largest crops in Lebanon. The research includes the creation of an innovative cooperative approach for crop classification from satellite images. The different crop maps created by different methods and approaches are verified using collected field samples. The final crop map which was created by the cooperative approach which includes the expert based classification model showed the highest accuracy of 86% compared to supervised and unsupervised classification algorithms such as Maximum Likelihood (ML), Minimum Distance (MD) and Iterative Self-Organizing Data Analysis Technique Algorithm (ISODATA) with accuracies of 34.8%, 44.1% and 14.2% respectively.
... Indices such as NDVI, SAVI and RVI have been used in such studies to show the relationship between grazers' distribution and forage production. This relationship between NDVI and vegetation phenology have been used in mapping and discriminating rangeland changes as could be observed on satellite image data (de Bie, Mobushir, Toxopeus, Venus, & Skidmore, 2008). However, the maps in most cases relied on a one time image data to map vegetation cover. ...
... Mapping techniques have varied from one researcher to another. De Bie et al. (2008) in their analysis of hyper-temporal images for crop mapping argue that many researchers have done land cover mapping by interpreting single time frame multi-spectral images leaving out of the map the aspect of high temporal variation, a major characteristic of vegetation. To effectively map and monitor grasslands, it is important to define map units of interests depending on the behaviour of vegetation in time as can be measured by the satellite sensors de Bie et al. (2008). ...
... De Bie et al. (2008) in their analysis of hyper-temporal images for crop mapping argue that many researchers have done land cover mapping by interpreting single time frame multi-spectral images leaving out of the map the aspect of high temporal variation, a major characteristic of vegetation. To effectively map and monitor grasslands, it is important to define map units of interests depending on the behaviour of vegetation in time as can be measured by the satellite sensors de Bie et al. (2008). Different vegetation types and species show different spectral profiles that are distinct and useful for discrimination and mapping. ...
Thesis
Full-text available
Rangeland vegetation mapping and assessment of its productivity is an integral aspect of ecosystem management. This study aims to map grasslands of Masai Mara ecosystem and estimate above ground grass biomass for rangeland monitoring and management. A review of previous vegetation maps show that there is need for a new mapping approach that solves the problem of misinterpretation of remote sensing data. Misinterpretation results from local distribution of rainfall which is highly variable in space and time in this area. Highly variable rainfall also affect rangeland seasonal productivity of forage in the rangeland. Therefore, a reliable model for estimating rangeland biomass that is not rainfall dependent is required. The methods used in mapping vegetation cover in this study involved; i) unsupervised image classification through ISODATA clustering and, ii) calculations of NDVI image stack statistics. Analysis of hyper-temporal Modis terra NDVI data produced classified NDVI and NDVI image statistics SD, Median and Trend. The image analysis outputs were used to design a sample scheme for fieldwork. Random stratified sampling was then followed to gather vegetation and biomass samples during fieldwork. Field samples were therefore analysed and used to characterize NDVI Classes into meaningful vegetation cover types. Biomass samples collected using quadrat and clipping technique were used to train biomass prediction model. Linear regression modelling technique was used to determine a statistically significant (p<0.05) model for predicting grass biomass. The statistical analysis also involved correlation coefficient calculation between measured grass biomass and explanatory variables SD, Median, Trend, distance to Bomas, animal density and NDVI as at Oct 2014. Root Mean Square Error (RMSE) was calculated for the model and used to assess its accuracy in prediction. The prediction model was validated using secondary data that was collected in Sept/Oct 2006 to check if the predicted values differ statistically to the 2006 measured data. The results of this study showed that it is possible to map vegetation cover through NDVI-derived data such as SD, Median and Trend. The mapping procedure distinguished the area into six cover units (A, B, C, D, E and F). However, some of the differences that are easily detected through remote sensing are not clearly distinguishable through field percent cover estimates because of overlaps in cover estimations. The differences in mapped cover types were investigated through a statistical test of difference using field measured grass biomass. A Kruskal-Wallis test reveal that mean of biomass measurements are significantly different between vegetation cover units; C – E, D – E, and E – F. Statistical results from Spearman’s Rank correlation tests revealed that grass biomass is significantly correlated to variables SD, distance to Bomas and to animal density. Linear relationship also exist between grass biomass and NDVI though not significant. Significant model coefficients explaining biomass (R2 = 0.653, N=42) was developed and used in predicting biomass. A Wilcoxon signed-rank test was done to compare between the model estimates and historical biomass measurements of 2006 and the results show that the two biomass datasets are not identical. This study concluded that the most reliable mapping approach to the effect of highly variable rainfall is through NDVI-Derived image products which measure the behaviour of vegetation over a longer period of time and not weather but climate dependent. However, it does not perform well in overlapping percentage cover estimates. This study have also demonstrated that SD, Bomas and NDVI measurements are key factors associated to measurable grass biomass and the approach used is not comparable to the one provided by IRLI, 2006 for this area since the results of the two studies are statistically significantly different. This study therefore recommends that, future studies should consider SD of NDVI more in vegetation cover mapping, assess biomass in different seasons with successive data and also include soil and herbivore grazing intensity in order to get an improved biomass prediction model.
... Land classification: We used the "separability" classification approach ( De Bie, Khan, Toxopeus, Venus, & Skidmore, 2008) ( Fig. 2). The ISODATA clustering algorithm (ErdasImagine software package) was run to set 19 different hypertemporal land classifications, each containing a different number of classes : 10, 20, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 40, 50 and 60. ...
... We, thus, suggest identifying class 4 with cloud semideciduous forests, whose occurrence at altitudes between 800 m and 1 500 m has also been described in the Guyana Shield (Huber, 1995a; 1995b). One of the main contributions of our landcover map is that every land class is individually related to a "time-profile-based signature" which forms part of the map legend (see De Bie et al., 2008). The pulsating patterns of hyper-temporal greenness profiles, which have defined every land-cover type, have led to us to label our map as "The Forest Pulse". ...
Article
Full-text available
Remote sensing and traditional ecological knowledge (TEK) can be combined to advance conservation of remote tropical regions, e.g. Amazonia, where intensive in situ surveys are often not possible. Integrating TEK into monitoring and management of these areas allows for community participation, as well as for offering novel insights into sustainable resource use. In this study, we developed a 250 m resolution land-cover map of the Western Guyana Shield (Venezuela) based on remote sensing, and used TEK to validate its relevance for indigenous livelihoods and land uses. We first employed a hyper-temporal remotely sensed vegetation index to derive a land classification system. During a 1 300 km, eight day fluvial expedition in roadless areas in the Amazonas State (Venezuela), we visited six indigenous communities who provided geo-referenced data on hunting, fishing and farming activities. We overlaid these TEK data onto the land classification map, to link land classes with indigenous use. We characterized land classes using patterns of greenness temporal change and topo-hydrological information, and proposed 12 land-cover types, grouped into five main landscapes: 1) water bodies; 2) open lands/forest edges; 3) evergreen forests; 4) submontane semideciduous forests, and 5) cloud forests. Each land cover class was identified with a pulsating profile describing temporal changes in greenness, hence we labelled our map as "The Forest Pulse". These greenness profiles showed a slightly increasing trend, for the period 2000 to 2009, in the land classes representing grassland and scrubland, and a slightly decreasing trend in the classes representing forests. This finding is consistent with a gain in carbon in grassland as a consequence of climate warming, and also with some loss of vegetation in the forests. Thus, our classification shows potential to assess future effects of climate change on landscape. Several classes were significantly connected with agriculture, fishing, overall hunting, and more specifically the hunting of primates, Mazama americana, Dasyprocta fuliginosa, and Tayassu pecari. Our results showed that TEK-based approaches can serve as a basis for validating the livelihood relevance of landscapes in high-value conservation areas, which can form the basis for furthering the management of natural resources in these regions.
... Change detection (CD) is the process of identifying changes in the state of an object by observing it at different time instances [56]. CD is important due to its application in various fields including disaster management [3,16,19,36,60], land use land cover (LULC) [2,12,32,43,45,46], urban expansion planning [6,23,40,44,57,65] crop monitoring [5] etc. During disaster management, CD helps in detecting regions undergoing massive upheavals during times of natural disasters through real time monitoring of the high-resolution satellite images. ...
Article
Full-text available
There is an increasing need to get updated information regarding the changes on earth’s surface. The information obtained can be used in a wide range of applications including disaster management, land-use investigation etc. The high-resolution remote sensing images obtained from satellites provide us with an opportunity to detect changes on earth’s surface between various time intervals. In this paper, an unsupervised object-based change detection (OBCD) method is proposed to detect changes in high resolution bi-temporal satellite images. To detect changes, a novel multi-feature non-seed-based region growing (MF-NSRG) algorithm is proposed for image segmentation based on heterogeneity minimization that uses textural heterogeneity along with spectral and spatial heterogeneity during region growing. The performance of MF-NSRG algorithm is further improved by using Harris Hawk, a recently proposed metaheuristic algorithm, which is used to obtain optimal values of segmentation parameters. Finally, the feature maps extracted from the pre-change and post-change segmented images are analysed using histogram trend similarity (HTS) approach to detect changes. The proposed approach is known as object-based change detection using Harris Hawk (OBCD-HH). The proposed OBCD-HH approach is applied on two datasets: xBD and Onera Satellite Change Detection (OSCD) dataset. Its performance is compared with existing state-of-the-art algorithms and results show the superiority of the proposed approach.
... Crop classification and analysis of crop dynamics are major tasks in agriculture-related applications. Remote sensing images assist in monitoring the crops 6,7 and their dynamics 8 throughout the year. In tropical countries such as India with a long coastal line, coconut production is an integral part of the country's economy. ...
Article
The coconut tree is an important cash crop in coastal tropical regions. These trees undergo substantial damage during annual cyclones. Early and effective assessment of damage is important to farmers for appropriate monetary compensation. However, the identification of coconut trees presents multiple challenges: (i) the trees are with other vegetation and (ii) cloud cover during monsoons and cyclones precludes using simple imaging techniques. Although automated approaches based on classical machine learning techniques have been attempted for other crops on remote sensed images, these are not adequate for identifying coconut trees from different land cover. We present a hybrid approach termed CocoNet that combines (i) unsupervised K-means for pixel-based coarse classification of land cover and (ii) a patch-based convolutional neural network for fine classification of vegetation to identify coconut farms. The challenge presented by cloud cover is addressed using synthetic aperture radar (SAR) images. Change detection is then performed on bitemporal Sentinel-1 SAR images taken before and after a cyclone to obtain the change map depicting the damage caused. Experimental results show that the proposed two-phase CocoNet model gives a classification accuracy of 92.5% and a change detection accuracy of 89.1%.
... The proposed approach is applied on areas of Malawi and Mozambique, where a small number of similar studies has been carried out so far, e.g. (De Bie et al., 2008;Gumma et al., 2019). With respect to the current state of the art, the contributions of this paper are: ...
Article
Full-text available
Crop type mapping currently represents an important problem in remote sensing. Accurate information on the extent and types of crops derived from remote sensing can help managing and improving agriculture especially for developing countries where such information is scarce. In this paper, high-resolution RGB drone images are the input data for the classification performed using a transfer learning (TL) approach. VGG16 and GoogLeNet, which are pre-trained convolutional neural networks (CNNs) used for classification tasks coming from computer vision, are considered for the mapping of the crop types. Thanks to the transferred knowledge, the proposed models can successfully classify the studied crop types with high overall accuracy for two considered cases, achieving up to almost 83% for the Malawi dataset and up to 90% for the Mozambique dataset. Notably, these results are comparable to the ones achieved by the same deep CNN architectures in many computer vision tasks. With regard to drone data analysis, application of deep CNN is very limited so far due to high requirements on the number of samples needed to train such complicated architectures. Our results demonstrate that the transfer learning is an efficient way to overcome this problem and take full advantage of the benefits of deep CNN architectures for drone-based crop type mapping. Moreover, based on experiments with different TL approaches we show that the number of frozen layers is an important parameter of TL and a fine-tuning of all the CNN weights results in significantly better performance than the approaches that apply fine-tuning only on some numbers of last layers.
... Climate variability is another factor contributing to the changing crop type due to the fact that over the years, crop yields have been reducing as a result of climate change translating to reduced economic returns and hence the need for farmers to increase their economic returns by planting introduced cash crops even though they are not adapted to the region. De bie et al. (2008) found that disparity between the crop types and the changing crop intensities were attributed to major droughts faced in India during the period of study. A similar study by Punithavathi, et al. (2012) to assess agricultural cropping concentration and crop wise changes, showed changes in crop types grown as a result of migration of people and poor climatic conditions due to climatic changes. ...
Article
Full-text available
Assessment of the distribution and dynamics of vegetation is becoming increasingly important in predicting the effects of climate change especially in the ASALs. It is therefore important to determine the effects of climate change on the crop type and land cover of the semi-arid regions hence the current study was carried out to assess the changes in crop type and land cover between 1986 – 2000 and 2000 – 2012 in Yatta District, Kenya. The LANDSAT TM, ETM and ETM+ satellite images of the years 1986, 2000 and 2012 in Yatta Sub-county were classified using ENVI 4.7 under supervised classification into different crop types and land cover. False colour composite using different reflective indexes (Bands 4, 3, 2) were used for the visual examination and interpretation of the images and maximum likelihood method of classification used. The percentage changes of crop types between 1986 – 2000 and 2000 – 2012 were determined using ENVI EX by comparing two images of different times. Questionnaires were administered to establish change detection from traditional (crops grown in the past but have been abandoned and underutilized) to introduced (crops grown as a result of technological advancement and economical advantage) crops in specific locations within the respondents farms. In 2012, maize and beans covered 72% while traditional crops (Sorghum, finger millet, cassava, dolichos, sweet potatoes, green grams, cowpeas, pigeon pea, and pumpkins), shrub land, bare land and riverine forest covered 14, 6, 3 and 5% of the study area, respectively. There was a significant (P=0.000) decline in the area under traditional crops (10.44 and 11.93 %), and a significant (P=0.000) increase in maize (4.70 and 22.73%), beans (23.83 and 2.6%) and bare land (3.42 and 1.03%) between the years 1986 – 2000 and 2000 – 2012 respectively. However, there was a significant (P=0.006) decrease in riverine vegetation (2.7 and 3.13 %) as well significant (P=0.000) decrease in shrub land (18.81 and 11.3 %) between the years 1986 – 2000 and 2000 – 2012 respectively. The observed trends will be important in guiding capacity builders on the crop type and land cover changes in the region who will in turn sensitize the community on the importance of traditional crops in view of the increasing climate variability and help in development of strategies for reintroduction of traditional crops in view of climate change and dwindling land resources as well as inform policies that will promote their reintroduction to achieve food security
... In fact, NDVI data have been succefully applied by different scientists for crops mapping, with ability to differentiate areas with different crop types (Liang, Bing-fang, Ji-hua et al., 2008;Wardlow, Egbert and Kastens, 2007). The data are rich in temporal dimension, hence, they are acknowledged for their ability to record vegetation dynamics of an area over time (de Bie, Khan, Toxopeus et al., 2008). ...
Article
Full-text available
Crop calendar is an important tool providing relevant information on crops cycles in a specific area for effective agricultural management. Crop calendars vary in different areas given dissimilarities in agro-ecosystems' characteristics. This research used multi-temporal MODIS NDVI stratification to assess differences in practiced maize crop calendars in various areas of Rwanda. Four (4) sample NDVI strata dominated by agriculture were purposively chosen, and 433 local farmers were randomly selected from the strata for interviews. The collected information helped to know about their maize planting as well as harvesting dates in order to generate maize calendars per NDVI strata. The generated crop calendars were later classified using k-means unsupervised classification, and produced 4 groupings of practiced maize calendars irrespective of NDVI strata. ANOVA results revealed significant differences between both the generated maize crop calendars by NDVI strata and the practiced crop calendars irrespective of NDVI strata, at p = 0.05. Moreover, chi-square tests and t-tests revealed not only a significant relationship between maize calendars and number of crop growing seasons, but also a significant relationship between maize calendars and NDVI strata, at p = 0.05. Finally, findings of this research contrasted the present conviction that there exist a single general maize calendar all over the country. Instead, the results were in accordance with the fact that Rwanda agro-ecosystems differ from East to West in terms of, mainly, altitude and rainfall patterns variations.
Article
Full-text available
False start of the growing season (Fsos) is a component of the onset variability related to agronomic drought that adversely impact on agricultural production and productivity. In the sub-Saharan Africa (SSA) where agriculture heavily depends on rainfall, the Fsos tends to create confusion among farmers on when to start planting crops thereby affecting seed germination and normal growth after emergence. In this paper, we focus on the Fsos and the occurrence of dry spell especially before the Start of growing Season (SoS). We take advantage of the existing rainfall estimates (CHIRPS) and remotely sensed data for vegetation performance (NDVI) over the period 1999-2017 in combination with local knowledge derived from farmers to map out areas at risk of (i) dry spell at the SoS, and (ii) false timing of SoS or high probability of occurrence of the Fsos. We found that the North Eastern part of Uganda (8.8% of arable area) were at risk of dry spell throughout each year. However, the greater North (58.1% of arable area) was prone to dry spell during the onset of the March-May season. Areas in the South Western (3.7%) region were at risk during the onset of the September-November season. The probability that a location in Uganda experiences an Fsos falls between 0-53%. The findings in this study are vital for planning of predictive adaptation to the impacts of climate variability on agriculture amid struggle aimed at tackling food insecurity challenge in the SSA.
Article
Landsat-based land-use land-cover (LULC) mapping studies were previously conducted in Giba catchment, comprising an area of 4019 km². No attempt has been done to map LULC of this catchment through the analysis of Moderate Resolution Imaging Spectroradiometer (MODIS) normalized difference vegetation index (NDVI) time-series data. This article is aimed to see whether time-series MODIS NDVI data set is applicable for LULC mapping of Giba catchment or not. MODIS NDVI data sets of the year 2010 were used for classification analysis. The original data were subjected to MODIS Reproduction Tool and stacking. The re-projected and stacked images were filtered using Harmonic Analysis of Time-Series filtering algorism to remove the effects of cloud and other noises. The MODIS NDVI data sets (16-day maximum value composite) were classified using the ISODATA clustering algorithm available under ERDAS IMAGINE software. A series of unsupervised classification runs were carried out with a pre-defined number of classes (5–24). From this classification, the optimal numbers of classes were determined to be eight after checking for average divergence analysis. The classification result became eight LULC classes namely: bare land, grass land, irrigated land, cultivated land, area closure, shrub land, bush land, and forest land with an overall accuracy of 87.7%. It was therefore concluded that MODIS NDVI time-series image is applicable for mapping large watersheds.
Article
Full-text available
Coarse spatial resolution, high temporal frequency satellite data from the NOAA/AVHRR system are presented to demonstrate their utility for monitoring vegetation seasonal dynamics. The techniques for processing and analysing the data are outlined and examples are given for selected applications at a range of scales. Normalized difference vegetation index images are presented for the entire globe and for the continents of Africa, South America and south-east Asia, with descriptions of the seasonal dynamics of major vegetation formations as portrayed by the transformed AVHRR data. Monitoring of forest clearance in Brazil, the productivity of African grasslands, Indian tropical forest and Chinese agriculture are selected for discussion. The paper concludes that coarse-resolution satellite data provide a valuable tool for vegetation mapping and monitoring at regional and global scales.
Article
Full-text available
Techniques based on multi-temporal, multi-spectral, satellite-sensor- acquired data have demonstrated potential as a means to detect, identify, map and monitor ecosystem changes, irrespective of their causal agents. This review paper, which summarizes the methods and the results of digital change detection in the optical/infrared domain, has as its primary objective a synthesis of the state of the art today. It approaches,digital change,detection from,three angles. First, the different perspectives from which the variability in ecosystems and the change,events have been dealt with are summarized.,Change,detection between pairs of images,(bi-temporal) as well as between,time profiles of imagery,derived indicators (temporal trajectories), and, where relevant, the appropriate choices for digital imagery acquisition timing and change interval length definition, are discussed. Second, pre-processing routines either to establish a more direct linkage between remote sensing data and biophysical phenomena, or to temporally mosaic imagery and extract time profiles, are reviewed. Third, the actual change,detection,methods,themselves,are categorized,in an analytical framework and critically evaluated. Ultimately, the paper highlights how some of these methodological,aspects are being,fine-tuned as this review,is being written, and we summarize the new developments that can be expected in the near future. The review,highlights the high complementarity,between,different change,detection methods.
Article
Data from the AVHRR sensor of polar-orbiting meteorological satellites were used to estimate biomass production on the southern border of the Sahara from 1981-1986. Average integrated normalized difference vegetation index (NDVI) threshold values were compared with several climatic zone, bioclimatic, agroclimatic, agroecological, and vegetation maps. Threshold index isolines agreed closely with major zone boundaries in several instances but also differed in some cases. -from Authors
Article
Drought is one of the major environmental disasters in southern Africa. In recent years, the damage from droughts to the environment and economies of some countries was extensive, and the death toll of livestock and wildlife was unprecedented. Weather data often come from a very sparse meteorological network, incomplete and/or not always available in good time to enable relatively accurate and timely large scale drought detection and monitoring. Therefore, data obtained from the Advanced Very High Resolution Radiometer (AVHRR) sensor on board the NOAA polar-orbiting satellites have been studied as a tool for drought monitoring and climate impact assessment in southern Africa. The AVHRR-based vegetation condition index (VCI) and temperature condition index (TCI) developed recently were used in this study because in other parts of the globe they showed good results when used for drought detection and tracking, monitoring excessive soil wetness, assessment of weather impacts on vegetation, and evaluation of vegetation health and productivity. The results clearly show that temporal and spatial characteristics of drought in southern Africa can be detected, tracked, and mapped by the VCI and TCI indices. These results were numerically validated by in situ data such as precipitation, atmospheric anomaly fields, and agricultural crop yield. In the later case, it was found that usable corn yield scenarios can be constructed from the VCI and TCI at approximately 6 (in some regions up to 13) weeks prior to harvest time. These indices can be especially beneficial when used together with ground data.
Article
In this paper we compare and contrast the performances of two low spatial resolution Earth observing systems, National Oceanographic and Atmospheric Administration - advanced very high resolution radiometer (NOAA-AVHRR) and Système pour l'Observation de la Terre - VEGETATION (SPOT-VGT), for burnt area mapping in a tropical savanna ecosystem of northern Australia. A multi-temporal composite image was derived from each of the two data sets acquired at the beginning of the 1999 fire season, and a set of rules for extracting burnt areas was compiled using a classification and regression trees (CART) methodology. In the case of the VGT data, the rules were used to segment the entire composite image; for the AVHRR data, a seed-growing approach was tested, with the aim of reducing overestimation of burnt areas. Since burnt areas are often confused with other land cover types, active fires were first identified (the seeds), and then the mapping of burnt areas was restricted to the neighbourhood of the pixels labelled as active fires. For the accuracy assessment, a reference map was derived by visual interpretation of two Landsat thematic mapper (TM) frames. First, the proportions of burnt areas were calculated for 14 × 14 km cells on a regular grid overlaid on AVHRR, VGT, and TM maps and the low-resolution areas were regressed against the TM estimates. Second, the spatial agreement between burnt area maps derived using high and low spatial resolution data was assessed using confusion matrices. We conclude that NOAA-AVHRR and SPOT-VGT show complementary characteristics. The better geometry of the VGT sensor, compared to the AVHRR, allows a higher accuracy in burnt area estimation, but the lack of a thermal channel on VGT is certainly a drawback for both the compositing and the burnt area mapping processes. Both systems tend to overestimate burnt areas when these areas are very fragmented but clearly underestimate when the burn scars are too small to be detected. Lastly, the seed-growing method is promising and could form the basis for a combined use of NOAA-AVHRR and SPOT-VGT systems.
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
A vegetation index (NDVI) time series of 1984 to 1993 was used to calibrate a pixel-based assessment of the start, end and length of the growing season in the sahelian country of Burkina Faso. Annual curves of rainfall/ETP, NDVI crop calendars were prepared for meteorological stations representative for three climatic zones. Agrometereological criteria for assessment of the start and the end were extracted from literature. NDVI parameters were assessed at the dekad of the agrometeorological onset and end of the growing season, and used for the definition of decision rules. Decision rules were implemented in a GIS to the NDVI image time series. The start of the growing season assessed with the NDVI dynamics was slightly better correlated with given farmer's sowing dates than with original agro-meteorological criteria. The normal start of the growing season, calculated from the NDVI image time series, was in good agreement with the recommended sowing dates. A delay of the onset of the growing season with regard to normal was monitored under near real time conditions in Burkina Faso in 1995. It was found that the delay of up to five dekads could lead to a partial crop failure and to changes in decisions of farmers. The model for the start and the end of the growing season was validated with independent data from different years. From the maps of the end and the start, maps of the length of the growing season were calculated by subtraction. The length of the growing season was found to be in agreement with agrometeorological literature, although in northern sahelian areas, the real growing season measured by the vegetation index tended to be shorter by about one dekad. In addition, the climatic map was improved by a per pixel classification of a normal (1984-1995 median) image time profile. The 36 normal NDVI images were classified using a hyperspectral image-processing tool.
Article
The researchers calculated seasonal integrated normalized difference vegetation index (NDVI) for each of 7 years using a time-series of 1-km data from the Advanced Very High Resolution Radiometer (AVHRR) (1992–93, 1995) and SPOT Vegetation (1998–2001) sensors. We used a local variance technique to identify each pixel as normal or either positively or negatively anomalous when compared to its surroundings. We then summarized the number of years that a given pixel was identified as an anomaly. The resulting anomaly maps were analysed using Landsat TM imagery and extensive ground knowledge to assess the results. This technique identified anomalies that can be linked to numerous anthropogenic impacts including agricultural and urban expansion, maintenance of protected areas and increased fallow. Local variance analysis is a reliable method for assessing vegetation degradation resulting from human pressures or increased land productivity from natural resource management practices.
Article
The dynamic nature of climate over Indian sub-continent is well known which influences Indian monsoon. Such dynamic variability of climate factors can also have significant implications for the vegetation and agricultural productivity of this region. Using empirical orthogonal function (EOF) and wavelet decomposition techniques, normalized difference vegetation index (NDVI) monthly data over Indian sub-continent for 18 years from 1982 to 2000 have been used to study the variability of vegetation. The present study shows that the monsoon precipitation and land surface temperature over the Indian sub-continent landmass have significant impact on the distribution of vegetation. Tropospheric aerosols exert a strong influence too, albeit secondary to monsoon precipitation and prove to be a powerful governing factor. Local climate anomaly is seen to be more effective in determining the vegetation change than any global teleconnection effects. The study documents the dominating influence of monsoon precipitation and highlights the importance of aerosols on the vegetation and necessitates the need for remedial measures. The present study and an earlier one point towards a possible global teleconnection pattern of ENSO as it is seen to affect a particular mode of vegetation worldwide.