Martin Gnyp

ABSTRACT This study assesses the use of TerraSAR-X data for monitoring rice cultivation in the Sanjiang Plain in Heilongjiang Province, Northeast China. The main objective is the understanding of the coherent co-polarized X-band backscattering signature of rice at different phenological stages in order to retrieve growth status.For this, multi-temporal dual polarimetric TerraSAR-X High Resolution SpotLight data (HH/VV) as well as single polarized StripMap (VV) data were acquired over the test site. In conjunction with the satellite data acquisition, a ground truth field campaign was carried out.The backscattering coefficients at HH and VV of the observed fields were extracted on the different dates and analysed as a function of rice phenology to provide a physical interpretation for the co-polarized backscatter response in a temporal and spatial manner. Then, a correlation analysis was carried out between TerraSAR-X backscattering signal and rice biomass of stem, leaf and head to evaluate the relationship with different vertical layers within the rice vegetation.HH and VV signatures show two phases of backscatter increase, one at the beginning up to 46 days after transplanting and a second one from 80 days after transplanting onwards. The first increase is related to increasing double bounce reflection from the surface–stem interaction. Then, a decreasing trend of both polarizations can be observed due to signal attenuation by increasing leaf density. A second slight increase is observed during senescence. Correlation analysis showed a significant relationship with different vertical layers at different phenological stages which prove the physical interpretation of X-band backscatter of rice. The seasonal backscatter coefficient showed that X-band is highly sensitive to changes in size, orientation and density of the dominant elements in the upper canopy.

The main objective of this study is to derive plant nitrogen (N) status and aboveground biomass via satellite remote sensing. To understand canopy spectral reflectance, the focus of the first part was set on the analysis of spectral signatures of winter wheat during its ...

ABSTRACT This paper illustrates the results obtained in the frame of experimental campaigns carried out on winter wheat fields in the North China Plain from March 2006 to June 2007. Investigations focused on the methodology of estimating biomass on a regional scale with hyperspectral (EO-1 Hyperion) and microwave data (Envisat ASAR). Special importance is drawn to the combined analysis of microwave and optical satellite data for crop monitoring. Since hyperspectral and synthetic aperture radar (SAR) sensors respond to crop characteristics differently, their complementary information content can support the estimation of crop conditions. During the regular field measurements, satellite data from jointing to ripening stages were acquired. Linear regression models between measured surface reflection as well as surface backscatter and wheat’s standing biomass were established. For hyperspectral data, the normalized ratio index (NRI) based on 825 nm and 1225 nm wavebands was calculated from 2006 data as input for the regression model. In addition, Envisat ASAR VV polarization data were related to winter wheat crop parameters. Bivariate correlation results from this study indicate that both multi-temporal EO-1 Hyperion as well as Envisat ASAR data provide notable relationships with crop conditions. As expected, linear correlation of hyperspectral data performed slightly better for biomass estimation (R2 = 0.83) than microwave data (R2 = 0.75) for the 2006 field survey. Based on the results, hyperspectral Hyperion data seem to be more sensitive to crop conditions. Improvements for crop parameter estimation were achieved by combining hyperspectral indices and microwave backscatter into a multiple regression analysis as a function of crop parameters. Combined analysis was performed for biomass estimation (R2 = 0.90) with notable improvements in prediction power.

Abstract: Timely monitoring of crop growth status at large scale is crucial for improving regional crop management decisions. The main objective of the recent study is a model development to predict and estimate crop parameters, here biomass, plant N ...

The working group for GIS and Remote Sensing at the Department of Geography at the University of Cologne has established a WebGIS called CampusGIS of the University of Cologne. The overall task of the CampusGIS is the connection of several existing databases at the University of Cologne with spatial data. These existing databases comprise data about staff, buildings, rooms, lectures, and general infrastructure like bus stops etc. These information were yet not linked to their spatial relation. Therefore, a GIS-based method is developed to link all the different databases to spatial entities. Due to the philosophy of the CampusGIS, an online-GUI is programmed which enables users to search for staff, buildings, or institutions. The query results are linked to the GIS database which allows the visualization of the spatial location of the searched entity. This system was established in 2005 and is operational since early 2006. In this contribution, the focus is on further developments. First results of (i) including routing services in, (ii) programming GUIs for mobile devices for, and (iii) including infrastructure management tools in the CampusGIS are presented. Consequently, the CampusGIS is not only available for spatial information retrieval and orientation. It also serves for on-campus navigation and administrative management.

ABSTRACT The non-destructive monitoring of crop growth status with field-based or tractor-based multi-or hyperspectral sensors is a common practice in precision agriculture. The demand is given for flexible, easy to use, and field scale systems with super-high resolution (< 20 cm) or on single plant scale to provide knowledge on in-field variability of crop status for management purposes. Satellite- and airborne systems are usually not able to provide the spatial and temporal resolution for such purposes within a low-cost approach. The developments in the area of Unmanned Aerial Vehicles (UAV) equipped with hyperspectral sensor systems may be suited to fill that niche. In this contribution, we introduce two hyperspectral full-frame cameras weighing less than 1 kg which can be mounted to low-weight UAVs (< 3 kg). The first results of a campaign in June/July 2013 are presented and the derived spectra from the hyperspectral images are compared to related spectra collected with a portable spectroradiometer. The first results are promising.

ABSTRACT In this study field experiments were conducted to test the ability of optimized spectral indices and partial least squares (PLS) to estimate leaf chlorophyll (Chl) content of rice from non-destructive canopy reflectance measurements. We integrated techniques involving the optimization of narrow band spectral indices and the detection of red edge position to optimize one type of spectral indices, the ratio of reflectance difference index (RRDI), for the estimation of leaf Chl content. The optimized RRDI in the red-edge (RRDIre = (R-745 - R-740)/(R-740-R-700)) accounted for 62% - 72% of the variation in leaf Chl content with an RMSE of 4.59 mu g/cm(2) - 4.89 mu g/cm(2). Compared to spectral indices, PLS improved the estimation of leaf Chl content, yielding R-2 and RMSE of 0.85 mu g/cm(2) and 3.22 mu g/cm(2), respectively. Finally, the model based on RRDI and the PLS model were further validated by an independent dataset collected in farmer fields. RRDI and PLS models yielded acceptable accuracy with R-2 of 0.49 and 0.55, respectively, and an RMSE of 5.47 mu g/cm(2) and 5.13 mu g/cm(2). Our results suggest the potential to optimize spectral indices and also the significance of PLS technique for mapping canopy biochemical variations.

ABSTRACT GreenSeeker active crop canopy sensor has been widely reported in literature for in-season nitrogen (N) management. An N fertilization optimization algorithm has been developed to use this sensor for topdressing N recommendation. An important component of this algorithm is to use crop sensor to estimate yield potential before topdressing N application. In addition to GreenSeeker, another commercial crop sensor, Crop Circle ACS - 470, has been recently developed with 6 possible spectral band choices (blue, green, two red bands, red edge and NIR). Little research has been conducted to compare these two sensors for estimating rice yield potential. The objective of this research was to determine how much improvements the Crop Circle ACS 470 would achieve for estimating rice yield potential at stem elongation stage, as compared with GreenSeeker sensor. Four N rate experiments were conducted in Jiansanjiang, Heilongjiang Province, in Northeast China in 2008 and 2009. The FieldSpec 3 hyperspectral canopy sensor was used to collect canopy reflectance, which was used to simulate spectral bands of GreenSeeker and Crop Circle sensors. The results indicated that in-season estimate of yield (INSEY) calculated with GreenSeeker NDVI and RVI could explain 42% and 62% of yield potential variability. The INSEY calculated with nine vegetation indices using simulated Crop Circle spectral bands performed better than GreenSeeker indices, with R2 being 0.64-0.75. INSEY (MTCARI) had the best performance, explaining 75% of the yield potential variability, followed by INSEY (MCARI1) (R2=0.73). The preliminary result indicated that Crop Circle ACS 470 sensor could explain 13% more variability in rice yield potential at stem elongation stage. More studies are needed to more systematically evaluate all possible vegetation indices that can be calculated with Crop Circle spectral bands for estimating rice yield potential at different growth stages.

ABSTRACT Hyperspectral vegetation indices have shown high potential for characterizing, classifying, monitoring, and modeling of vegetation and agricultural crops. Correlation matrices from hyperspectral vegetation indices and plant growth parameters help select important wavelength domains and identify redundant bands. We introduce the software HyperCor for automated pre-processing of narrowband hyperspectral field data and computation of correlation matrices. In addition, we propose a multi-correlation matrix strategy which combines multiple correlation matrices from different datasets and uses more information from each matrix. We apply this method to a large multi-temporal spectral library to derive vegetation indices and related regression models for rice biomass detection in the tillering, stem elongation, heading and across all growth stages. The models are calibrated with data from three consecutive years and validated with two other years. The results reveal that the multi-correlation matrix strategy can improve the model performance by 10 to 62 percent, depending on the growth stage.

ABSTRACT This paper contributes an assessment for estimating rice (Oryza sativa L., irrigated lowland rice) biomass by canopy re%ectance in the Sanjiang Plain, China. Hyperspectral data were captured with 'eld spectroradiometers in experimental 'eld plots and farmers’ 'elds and then accompanied by destructive aboveground biomass (AGB) sampling at different phenological growth stages. Best single bands, best two band-combinations, optimised simple ratio (SR), and optimised normalized ratio index (NRI), as well as multiple linear regression (MLR) were calculated from the re%ectance for the non-destructive estimation of rice AGB. Experimental 'eld data were used as the calibration dataset and farmers’ 'eld data as the validation dataset. Re%ectance analyses display several sensitive bands correlated to rice AGB, e.g. 550, 670, 708, 936, 1125, and 1670 nm, which changed depending on the phenological growth stages. These bands were detected by correlograms for SR, NRI, and MLR with an offset of approximately ± 10 nm. The assessment of the three methods showed clear advantages of MLR over SR and NRI in estimating rice AGB at the tillering and stem elongation stages by 'tting and evaluating the models. The optimal band number forMLR was set to four to avoid over'tting. The best validatedMLR model (R2 = 0.82) at the tillering stage was using four bands at 672, 696, 814 and 707 nm. Overall, the optimized SR, NRI, and MLR have a great potential in non-destructive estimation of rice AGB at different phenological stages. The performance against the validation dataset showed R2 of 0.69 for SR and R2 of 0.70 for NRI, respectively

ABSTRACT The influence of morphophysiological variation at different growth stages on the performance of vegetation indices for estimating plant N status has been confirmed. However, the underlying mechanisms explaining how this variation impacts hyperspectral measures and canopy N status are poorly understood. In this study, four field experiments involving different N rates were conducted to optimize the selection of sensitive bands and evaluate their performance for modeling canopy N status of rice at various growth stages in 2007 and 2008. The results indicate that growth stages negatively affect hyperspectral indices in different ways in modeling leaf N concentration (LNC), plant N concentration (PNC) and plant N uptake (PNU). Published hyperspectral indices showed serious limitations in estimating LNC, PNC and PNU. The newly proposed best 2-band indices significantly improved the accuracy for modeling PNU (R2 = 0.75–0.85) by using the lambda by lambda band-optimized algorithm. However, the newly proposed 2-band indices still have limitations in modeling LNC and PNC because the use of only 2-band indices is not fully adequate to provide the maximum N-related information. The optimum multiple narrow band reflectance (OMNBR) models significantly increase the accuracy for estimating the LNC (R2 = 0.67–0.71) and PNC (R2 = 0.57–0.78) with six bands. Results suggest the combinations of center of red-edge (735 nm) with longer red-edge bands (730–760 nm) are very efficient for estimating PNC after heading, whereas the combinations of blue with green bands are more efficient for modeling PNC across all stages. The center of red-edge (730–735 nm) paired with early NIR bands (775–808 nm) are predominant in estimating PNU before heading, whereas the longer red-edge (750 nm) paired with the center of “NIR shoulder” (840–850 nm) are dominant in estimating PNU after heading and across all stages. The OMNBR models have the advantage of modeling canopy N status for the entire growth period. However, the best 2-band indices are much easier to use. Alternatively, it is also possible to use the best 2-band indices to monitor PNU before heading and PNC after heading. This study systematically explains the influences of N dilution effect on hyperspectral band combinations in relating to the different N variables and further recommends the best band combinations which may provide an insight for developing new hyperspectral vegetation indices.

ABSTRACT Traditional vegetation indices like normalized difference vegetation index (NDVI) tend to saturate under high canopy coverage conditions and do not perform very well for diagnosing nitrogen status of high yielding crops. Red edge indices have been shown to be more sensitive to nitrogen status and are promising alternatives. The objective of this study is to evaluate different red edge indices for estimating winter wheat (Triticum aestivum L.) nitrogen status under high canopy coverage conditions in farmers' fields. Six 200 × 200 m fields in four villages from Shandong Province, China were selected for this research from 2005 to 2007. Hyperspectral reflectance was determined at Feekes growth stage 7-9 when wheat canopy was fully covered. Subsequently, linear relationships between red edge indices and plant N uptake and concentration were established. Results indicated that red edge position indices REP and REIP performed better in estimating plant N status than normalized and simple ratio red edge indices. Plant N uptake, as a better indicator of crop N status, was easier to estimate using red edge indices than plant N concentration.