Use of remotely sensed and ancillary data for estimating forest gross primary productivity in Italy

The current paper describes the development and testing of a procedure which can use widely available remotely sensed and ancillary data to assess large-scale patterns of forest productivity in Italy. To reach this objective a straightforward model (C-Fix) was applied which is based on the relationship between photosynthetically active radiation absorbed by plant canopies and relevant gross primary productivity (GPP). The original C-Fix methodology was improved by using more abundant ancillary information and more efficient techniques for NDVI data processing. In particular, two extraction methods were applied to NDVI data, derived from two sensors (NOAA-AVHRR and SPOT-VGT) to feed C-Fix. The accuracy of the model outputs was assessed through comparison with annual and monthly values of forest GPP derived from eight eddy covariance flux towers. The results obtained indicated the superiority of SPOT-VGT over NOAA-AVHRR data and a higher efficiency of the more advanced NDVI extraction method. Globally, the procedure was proved to be of easy and objective implementation and allowed the evaluation of mean productivity levels of existing forests on the national scale.     

Estimation of total above-ground phytomass production using remotely sensed data

Remote sensing potentially offers a quick and nondestructive method for monitoring plant canopy condition and development. In this study, multispectral reflectance and thermal emittance data were used in conjunction with micrometeorological data in a simple model to estimate above-ground total dry phytomass production of several spring wheat canopies. The fraction of absorbed photosynthetic radiation (PAR) by plants was estimated from measurements of visible and near-infrared canopy reflectance. Canopy radiation temperature was used as a crop stress indicator in the model. Estimated above-ground phytomass values based on this model were strongly correlated with the measured phytomass values for a wide range of climate and plant-canopy conditions.     

Estimation of net primary productivity by integrating remote sensing data with an ecosystem model

This paper describes a method for integrating leaf area index (LAI) derived from remote sensing data with an ecosystem model for accurate estimation of net primary productivity (NPP). The ecosystem model used in this study was Sim-CYCLE, with which LAI retrieved from the data acquired by MODIS sensor (MODIS-LAI) was integrated. Global annual NPP was estimated as 59.6 Gt C year⁻¹ by MOD-Sim-CYCLE (Sim-CYCLE after integration of MODIS-LAI), whereas it was 62.7 Gt C year⁻¹ in case of Sim-CYCLE for the year 2001. Both models predicted highest NPP around the equator with another smaller peak occurring around 60°N. These two regions represented the tropical and boreal forests biomes, respectively. The NPP estimated by MOD-Sim-CYCLE exceeded the NPP estimated by Sim-CYCLE in these two regions. Other than the tropical and boreal forests biomes, NPP values estimated by the MOD-Sim-CYCLE were typically lower than Sim-CYCLE across the latitudes. Validations of results in Australia and USA showed that MOD-Sim-CYCLE estimated NPP more accurately than Sim-CYCLE. Our results demonstrate the utility of combining satellite-observation with an ecosystem process model to achieve improved accuracy in estimates and monitoring global net primary productivity.     

Aboveground biomass estimation in intervened and non-intervened Nothofagus pumilio forests using remotely sensed data

Forest inventory data can be used along with remotely sensed data to estimate biomass and carbon stocks over large and inaccessible forested areas. In this study, the relationship between satellite-derived multispectral data and forest variables from intervened and non-intervened Nothofagus pumilio forest stands located in the Magellan region of Chile was examined, in order to quantify the over bark volume (OBV) and aboveground tree biomass (AGTB). Four vegetation parameters – the green normalised difference vegetation index (GNDVI), normalised difference vegetation index (NDVI), simple ratio (SR) and vegetation cover fraction (VCF) – were retrieved from an Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) image of the study area. The results indicate that only the VCF presents significant differences among intervened and non-intervened stands. The best OBV and AGTB models (R² = 0.58) were found using the SR index and the VCF as predictors. This result could be transferred to estimate biomass and volume in other Nothofagus pumilio forests with similar conditions. Moreover, it can be used to assess temporal carbon changes.     

Crop yield prediction from remotely sensed vegetation indices and primary productivity in arid and semi-arid lands

Global demands for biomass and arable lands are expected to double in the next 35 years. Scarcity of water resources in arid and semi-arid areas poses a serious threat to their agricultural productivity and hence their food security. In this study, we examine whether crop yields can be predicted from remotely sensed vegetation indices and remotely sensed estimates of primary productivity. Spatial relationships between remotely sensed enhanced vegetation index (EVI), net photosynthesis (P_Net), and gross and net primary production (GPP and NPP, respectively) in irrigated semi-arid and arid agro-ecosystems since the beginning of the century are analysed. The conflict-affected country of Syria is selected as the case study. Relationships between EVI and crop yield are investigated in an effort to enhance food production estimates in affected areas outside governmental jurisdictions. Estimates of NPP derived from reported irrigated agriculture crop data in a semi-arid and an arid zone are compared to remotely sensed NPP in a geospatial environment. Results show that winter crop yields are correlated with spring GPP in semi-arid zones of the study area (R² = 0.85). Summer crop yield can be predicted from either cumulative summer EVI (R² = 0.77) or P_Net in most zones. Where fully irrigated fields are surrounded by hyper-arid landscape, summer P_Net was negative in all instances and EVI was inversely correlated with yield. NPP from crops was much higher (290 gC m^?2 year^?1) in those regions than MOD17 NPP (70 gC m^–2), where 1.0 g of carbon is equivalent to 2.2 g of oven-dry organic matter (= 45% carbon by weight). The gap was less in semi-arid zones (2–39% difference). Overall crop-derived NPP for the period 2000–2013 was 322 versus 300 gC m^–2 for that remotely sensed within the cropped zones of the political units. The results of this study are crucial to derive accurate estimates of irrigated agriculture productivity and to study the effect of the latter on net ecosystem carbon storage.     

Calibrated tree counting on remotely sensed images of planted forests

The development of robust and accurate methods for counting trees from remotely sensed data could provide substantial cost savings in forest inventory. A new methodology that provides a framework for calibrating tree detection algorithms to obtain accurate tree counts for even-aged stands is described. The methodology was evaluated using two tree detection algorithms and two operators using airborne laser scanning (ALS) and orthophotograph images for four Pinus radiata D.Don stands ranging in age between 5 and 32 years with stand densities ranging between 204 and 826 stems ha^?1. For application of the methodology to ALS images the error of estimate on the total count was 4.7% when calibration counts from actual ground plots were used and 10.5% when calibration counts from virtual plots on the image were used. For orthophotographs, the error of estimate was 6.1% using ground calibration plots and 24.3% using calibration counts from virtual plots. The described methodology was shown to be robust to variations in the process from the two operators and two algorithms evaluated. The measure of accuracy determined using the methodology can be used to provide an objective basis for evaluating a wide range of tree counting and detection processes in future research.     

Estimation of leaf water status to monitor the risk of forest fires by using remotely sensed data

Estimating the water status of vegetation is one of the most important elements in assessing forest fire danger. In this paper, laboratory measurement confirmed a relationship between leaf water status and the normalized difference water index (NDWI), derived from near-infrared and shortwave-infrared spectral data. Two results were confirmed: (a) NDWI is related to equivalent water thickness, and, (b) in addition to NDWI, the quantity of leaf material must be known in order to estimate vegetation dryness. Based on these findings, the authors developed a vegetation dryness index (VDI) to estimate global vegetation water content. VDI values, calculated by using SPOT/VEGETATION data, were applied to data from a 1998 forest fire in the Russian Far East. This led to two results: (a) VDI was useful for detecting areas with a high potential for ignition, and (b) VDI may have been able to detect the fire-spread direction.     

Burn severity estimation from remotely sensed data: Performance of simulation versus empirical models

Burn severity is a key factor in post-fire assessment and its estimation is traditionally restricted to field work and empirical fitting from remotely sensed data. However, the first method is limited in terms of spatial coverage and cost effectiveness and the second is site- and data-specific. Since alternative approaches based on radiative transfer models (RTM) have been usefully applied in retrieving several biophysical plant parameters (leaf area index, water and dry matter content, chlorophyll), this paper has applied the inversion of a simulation model to estimate burn severity in terms of the Composite Burn Index (CBI). The performance of the model inversion method was compared to standard empirical techniques. The study area chosen was a large forest fire in central Spain which occurred in July 2005. The model inversion showed the most accurate estimation for high severity levels (for CBI > 2.7, RMSE = 0.30) and for unburned areas (CBI < 0.5, RMSE = 0). In both methodologies, the error associated to CBI from 0.5 to 2.7 was not acceptable (RMSE > 0.7), because it is higher than 25% of the total range of the index. Finally, burn severity maps from both methods were compared.     

An approach to linking remotely sensed data and areal census data

With the increasing availability of remotely sensed data and census data, discussing their relationship is one of the important issues in GIS data integration. This paper proposed an approach to linking three levels (macro, medium and micro) of land classifications with areal census data on hierarchical census boundaries. Specifically, a method of building the correlations between areal census dwelling data and residential densities classified by a remote sensing approach was demonstrated. First, a texture statistic (homogeneity) along with six Thematic Mapper (TM) bands (bands 1-5 and 7) was put together to classify residential density levels. The homogeneity slightly enhanced classification accuracy. Then, to test the correlations between census dwelling data and residential densities, a multiple linear regression was conducted. It was found that areal census dwelling data had higher correlations with areas of different residential densities than with the aggregated area of a whole residential area at an individual census zone level. Finally, the paper discussed that dis-aggregation of areal census data based on dwelling densities within the framework of remote sensing and GIS would be very useful for multidisciplinary studies, such as natural hazards risk assessment.     

Improving specific class mapping from remotely sensed data by cost-sensitive learning

In many remote-sensing projects, one is usually interested in a small number of land-cover classes present in a study area and not in all the land-cover classes that make-up the landscape. Previous studies in supervised classification of satellite images have tackled specific class mapping problem by isolating the classes of interest and combining all other classes into one large class, usually called others, and by developing a binary classifier to discriminate the class of interest from the others. Here, this approach is called focused approach. The strength of the focused approach is to decompose the original multi-class supervised classification problem into a binary classification problem, focusing the process on the discrimination of the class of interest. Previous studies have shown that this method is able to discriminate more accurately the classes of interest when compared with the standard multi-class supervised approach. However, it may be susceptible to data imbalance problems present in the training data set, since the classes of interest are often a small part of the training set. A result the classification may be biased towards the largest classes and, thus, be sub-optimal for the discrimination of the classes of interest. This study presents a way to minimize the effects of data imbalance problems in specific class mapping using cost-sensitive learning. In this approach errors committed in the minority class are treated as being costlier than errors committed in the majority class. Cost-sensitive approaches are typically implemented by weighting training data points accordingly to their importance to the analysis. By changing the weight of individual data points, it is possible to shift the weight from the larger classes to the smaller ones, balancing the data set. To illustrate the use of the cost-sensitive approach to map specific classes of interest, a series of experiments with weighted support vector machines classifier and Landsat Thematic Mapper data were conducted to discriminate two types of mangrove forest (high-mangrove and low-mangrove) in Saloum estuary, Senegal, a United Nations Educational, Scientific and Cultural Organisation World Heritage site. Results suggest an increase in overall classification accuracy with the use of cost-sensitive method (97.3%) over the standard multi-class (94.3%) and the focused approach (91.0%). In particular, cost-sensitive method yielded higher sensitivity and specificity values on the discrimination of the classes of interest when compared with the standard multi-class and focused approaches.     

Image processing for boundary extraction of remotely sensed data

Image processing algorithms for analysing remotely sensed data are developed. The algorithms proposed in the paper provide means for autoregressive texture modelling and for boundary detection of uniform subimage areas. The boundary detection methods make use of a semicircular entropy operator and of the binary hypothesis testing of the Poisson counting process. The proposed algorithms are applied to the pattern analysis of the isothermal distribution in the oceanic environment.     

Modelling net primary productivity of terrestrial ecosystems in East Asia based on an improved CASA ecosystem model

By using a land cover map, normalized difference vegetation index (NDVI) data sets, monthly meteorological data and observed net primary productivity (NPP) data, we have improved the method of estimating light use efficiency (LUE) for different biomes and soil moisture coefficients in the Carnegie–Ames–Stanford Approach (CASA) ecosystem model. Based on this improved model we produced an annual NPP map (in 1999) for the East Asia region located at 10–70° N, 70–170° E (about 19.66% of the terrestrial surface of the Earth). The results show that the mean NPP for the study area in 1999 was 374.12 g carbon (C) m^?2 year^?1 and the total NPP was 1.096 × 10¹⁴ kg C year^?1, making up 17.51–18.39% of the global NPP. Comparison between the estimated NPP obtained from this improved CASA ecosystem model and the observed NPP obtained from two NPP databases indicates that the estimated NPP is close to the observed NPP, with an average error of 5.15% for the study region. We used two different land cover maps of China to drive the improved CASA model by keeping other inputs unchanged to determine how the classification accuracy of the land cover map affects the estimated NPP, and the results indicate that an accurate land cover map is important for obtaining an accurate and reliable estimate of NPP for some regions, especially for a particular biome.     

Estimation of the global net primary productivity using NOAA images and meteorological data: changes between 1988 and 1993

A model for net primary productivity (NPP) estimation was developed based on a relationship between NPP estimated by the Chikugo model and the intensity-sum of the normalized difference vegetation index (NDVI) multiplied by the solar radiation during growth periods. There was a clear linear relationship between the estimated NPP and the intensity-sum (R ²=0.845), whose slope indicated the average light use efficiency (LUE) of global plants. The NPP estimation model (NDVI-based model), which included growth multipliers of optimum air temperature and soil water stress on vegetation growth with LUE, was developed. NDVI anomalies caused by scattering of volcanic ash from Mt Pinatubo were reduced by a correction based on intensity matching of channels 1 and 2 individually. NDVI retrieved a seasonal change pattern in 1991 and 1992 after the correction. Global NPP between 1988 and 1993 was estimated using the NDVI-based model, corrected NDVI, air temperature and soil water content data. There was a linear relationship between the estimated NPP and NPP observed in forests in China. The average global NPP during the 6 years was about 123?Pg dry weight per year, and the maximum and minimum NPP appeared in 1991 and 1988, respectively.     

Estimation of winter wheat acreage via a combination of remotely sensed data and an optimized spatial sampling scheme

Timely and accurate estimates of crop areas are critical to enhancing agriculture management and ensuring national food security. This study aims to combine remote-sensing data and an optimized spatial sampling scheme to improve the accuracy of crop area estimates and decrease the cost of crop surveys at a regional scale. This study focuses on winter wheat in Mengcheng County in Anhui Province, China. Advanced Land Observing Satellite (ALOS) Advanced Visible light and Near Infrared Radiometer (AVNIR)-2, and Landsat5 Thematic Mapper (TM) images from 2009 and 2010, respectively, are used to extract the winter wheat area and distribution. Additionally, a spatial sampling scheme was optimized by combining remotely sensed data, geographical information system (GIS), Geostatistics, and traditional sampling methods. The experimental results demonstrate that the variability in the proportion of winter wheat acreage in one sampling unit (PWS) increases with increasing sampling unit size. The PWS coefficient of variation (CV) varies from 32.75 to 43.46% among the eight sampling unit sizes. The spatial correlation thresholds of PWS increase with increasing sampling unit size. For small sampling unit sizes (500 m × 500 m–2000 m × 2000 m), the relative error and CV of the population extrapolation for the optimized sample layout are obviously lower than those of the simple random sampling method. For larger sampling unit sizes (2500 m × 2500 m–4000 m × 4000 m), the sample size is obviously lower for the optimized sample layout compared with that of the simple random sampling method, but there are no differences in the relative errors or CVs. By combining remote-sensing data and the optimized spatial sampling scheme, this research can improve the accuracy of crop area estimation at a regional scale.     

Supervised classification of remotely sensed data with ongoing learning capability

A methodology to implement an automatic system for classifying remotely sensed data with an ongoing learning capability is introduced. The Nearest Neighbour (NN) rule is employed as the central classifier and several techniques are added to cope with the increase in computational load and with the risk of incorporating noisy data into the training sample. Experimental results confirm the enhancement in classification accuracy.     

Statistical evaluation of remotely sensed data for water quality monitoring

The primary objective of this study was to determine relationships between water quality parameters (WQPs) and digital data from the Landsat satellite to estimate and map the WQP in the Porsuk Dam reservoir. Suspended sediments (SS), chlorophyll a (chl-a), NO₃-N and transmitted light intensity depth (TLID) were the parameters for water quality determination used in this study. Collection of these data, obtained from the General Directorate of State Hydraulic Works (GDSHW) was synchronized with the Landsat satellite overpass of the September 1987. The relationships between the brightness values (BV) of the TM data and WQP were determined. Using the TM data, we developed multiple regression equations to estimate the WQPs, and the validation of these equations was checked by using ANOVA. The effects of SS, NO₃-N and chl-a on TLID were tested not only for ground data, but also for TM datasets. Regression equations were developed for two different datasets and the homogeneity of those equations was tested. Finally, these regression equations evaluated from digital TM data and ground data were applied to map TLID values.     

Symbolic agglomerative clustering for quantitative analysis of remotely sensed data

An efficient nonparametric, hierarchical, symbolic agglomerative clustering procedure based on the mutual nearest neighbourhood concept is proposed for classifying remotely sensed multispectral data. The procedure utilized a data reduction technique and an innovative symbolic concept to minimize the memory and computational time requirements. A new non-metric similarity measure and a novel method of formulation of composite symbolic objects are proposed to enrich the performance of the algorithm. A Mean Difference Index (MDI) concept for identifying the optimal number of classes was used. Experiments were conducted on IRS (Indian Remote Sensing) satellite data to authenticate the efficacy of the procedure.     

Automatic land cover analysis for Tenerife by supervised classification using remotely sensed data

Automatic land cover classification from satellite images is an important topic in many remote sensing applications. In this paper, we consider three different statistical approaches to tackle this problem: two of them, namely the well-known maximum likelihood classification (ML) and the support vector machine (SVM), are noncontextual methods. The third one, iterated conditional modes (ICM), exploits spatial context by using a Markov random field. We apply these methods to Landsat 5 Thematic Mapper (TM) data from Tenerife, the largest of the Canary Islands. Due to the size and the strong relief of the island, ground truth data could be collected only sparsely by examination of test areas for previously defined land cover classes.We show that after application of an unsupervised clustering method to identify subclasses, all classification algorithms give satisfactory results (with statistical overall accuracy of about 90%) if the model parameters are selected appropriately. Although being superior to ML theoretically, both SVM and ICM have to be used carefully: ICM is able to improve ML, but when applied for too many iterations, spatially small sample areas are smoothed away, leading to statistically slightly worse classification results. SVM yields better statistical results than ML, but when investigated visually, the classification result is not completely satisfying. This is due to the fact that no a priori information on the frequency of occurrence of a class was used in this context, which helps ML to limit the unlikely classes.     

Forest biomass estimation through NDVI composites. The role of remotely sensed data to assess Spanish forests as carbon sinks

The present study is focused on the capabilities of remote sensing data and techniques to help in the monitoring of forest ecosystems as carbon sinks. It will attempt to find statistical relationships between satellite‐derived NDVI (Normalized Difference Vegetation Index) data from SPOT‐VEGETATION and NOAA‐AVHRR and field measurements from the Spanish National Forest Inventory on the geographical basis of provinces. Statistically significant relationships were obtained when correlating the aforementioned datasets. These relationships were then used to predict forest biomass at a national level, in order to obtain updated forest information between consecutive National Forest Inventories.