Bastian Siegmann

Comparison of different regression models and validation techniques for the assessment of wheat leaf area index from hyperspectral data

Leaf area index (LAI) is one of the most important plant parameters when observing agricultural crops and a decisive factor for yield estimates. Remote-sensing data provide spectral information on large areas and allow for a detailed quantitative assessment of LAI and other plant parameters. The present study compared support vector regression (SVR), random forest regression (RFR), and partial least-squares regression (PLSR) and their achieved model qualities for the assessment of LAI from wheat reflectance spectra. In this context, the validation technique used for verifying the accuracy of an empirical–statistical regression model was very important in order to allow the spatial transferability of models to unknown data. Thus, two different validation methods, leave-one-out cross-validation (cv) and independent validation (iv), were performed to determine model accuracy. The LAI and field reflectance spectra of 124 plots were collected from four fields during two stages of plant development in 2011 and 2012. In the case of cross-validation for the separate years, as well as the entire data set, SVR provided the best results (2011: R2cv = 0.739, 2012: R2cv = 0.85, 2011 and 2012: R2cv = 0.944). Independent validation of the data set from both years led to completely different results. The accuracy of PLSR (R2iv = 0.912) and RFR (R2iv = 0.770) remained almost at the same level as that of cross-validation, while SVR showed a clear decline in model performance (R2iv = 0.769). The results indicate that regression model robustness largely depends on the applied validation approach and the data range of the LAI used for model building.

An Overview of the Regional Experiments for Land-atmosphere Exchanges 2012 (REFLEX 2012) Campaign

The REFLEX 2012 campaign was initiated as part of a training course on the organization of an airborne campaign to support advancement of the understanding of land-atmosphere interaction processes. This article describes the campaign, its objectives and observations, remote as well as in situ. The observations took place at the experimental Las Tiesas farm in an agricultural area in the south of Spain. During the period of ten days, measurements were made to capture the main processes controlling the local and regional land-atmosphere exchanges. Apart from multi-temporal, multi-directional and multi-spatial space-borne and airborne observations, measurements of the local meteorology, energy fluxes, soil temperature profiles, soil moisture profiles, surface temperature, canopy structure as well as leaf-level measurements were carried out. Additional thermo-dynamical monitoring took place at selected sites. After presenting the different types of measurements, some examples are given to illustrate the potential of the observations made.

The Potential of Pan-Sharpened EnMAP Data for the Assessment of Wheat LAI

In modern agriculture, the spatially differentiated assessment of the leaf area index (LAI) is of utmost importance to allow an adapted field management. Current hyperspectral satellite systems provide information with a high spectral but only a medium spatial resolution. Due to the limited ground sampling distance (GSD), hyperspectral satellite images are often insufficient for precision agricultural applications. In the presented study, simulated hyperspectral data of the upcoming Environmental Mapping and Analysis Program (EnMAP) mission (30 m GSD) covering an agricultural region were pan-sharpened with higher resolution panchromatic aisaEAGLE (airborne imaging spectrometer for applications EAGLE) (3 m GSD) and simulated Sentinel-2 images (10 m GSD) using the spectral preserving Ehlers Fusion. As fusion evaluation criteria, the spectral angle (αspec) and the correlation coefficient (R) were calculated to determine the spectral preservation capability of the fusion results. Additionally, partial least squares regression (PLSR) models were built based on the EnMAP images, the fused datasets and the original aisaEAGLE hyperspectral data to spatially predict the LAI of two wheat fields. The aisaEAGLE model provided the best results (R2cv = 0.87) followed by the models built with the fused datasets (EnMAP–aisaEAGLE and EnMAP–Sentinel-2 fusion each with a R2cv of 0.75) and the simulated EnMAP data (R2cv = 0.68). The results showed the suitability of pan-sharpened EnMAP data for a reliable spatial prediction of LAI and underlined the potential of pan-sharpening to enhance spatial resolution as required for precision agriculture applications.

Regionalization of Uncovered Agricultural Soils Based on Organic Carbon and Soil Texture Estimations

The determination of soil texture and organic carbon across agricultural areas provides important information to derive soil condition. Precise digital soil maps can help to till agricultural fields with more accuracy, greater cost-efficiency and better environmental protection. In the present study, the laboratory analysis of sand, silt, clay and soil organic carbon (SOC) content was combined with hyperspectral image data to estimate the distribution of soil texture and SOC across an agricultural area. The aim was to identify regions with similar soil properties and derive uniform soil regions based on this information. Soil parameter data and corresponding laboratory spectra were used to calibrate cross-validated (leave-one-out) partial least squares regression (PLSR) models, resulting in robust models for sand (R2 = 0.77, root-mean-square error (RMSE) = 5.37) and SOC (R2 = 0.89, RMSE = 0.27), as well as moderate models for silt (R2 = 0.62, RMSE = 5.46) and clay (R2 = 0.53, RMSE = 2.39). The regression models were applied to Airborne Imaging Spectrometer for Applications DUAL (aisaDUAL) hyperspectral image data to spatially estimate the concentration of these parameters. Afterwards, a decision tree, based on the Food and Agriculture Organization (FAO) soil texture classification scheme, was developed to determine the soil texture for each pixel of the hyperspectral airborne data. These soil texture regions were further refined with the spatial SOC estimations. The developed method is useful to identify spatial regions with similar soil properties, which can provide a vital information source for an adapted treatment of agricultural fields in terms of the necessary amount of fertilizers or water. The approach can also be adapted to wider regions with a larger sample size to create detailed digital soil maps (DSMs). Further, the presented method should be applied to future hyperspectral satellite missions like Environmental Mapping and Analysis Program (EnMap) and Hyperspectral Infrared Imager (HyspIRI) to cover larger areas in shorter time intervals. Updated DSMs on a regular basis could particularly support precision farming aspects.

On the potential of Wireless Sensor Networks for the in-situ assessment of crop leaf area index

Design of a novel low-cost sensor modification for non-destructive LAI assessment.Maize field campaigns including a comparative analysis with a standard instrument.An impact evaluation showing high accuracy and robustness of our approach. A precise and continuous in-situ monitoring of bio-physical crop parameters is crucial for the efficiency and sustainability in modern agriculture. The leaf area index (LAI) is an important key parameter allowing to derive vital crop information. As it serves as a valuable indicator for yield-limiting processes, it contributes to situational awareness ranging from agricultural optimization to global economy. This paper presents a feasible, robust, and low-cost modification of commercial off-the-shelf photosynthetically active radiation (PAR) sensors, which significantly enhances the potential of Wireless Sensor Network (WSN) technology for non-destructive in-situ LAI assessment. In order to minimize environmental influences such as direct solar radiation and scattering effects, we upgrade such a sensor with a specific diffuser combined with an appropriate optical band-pass filter. We propose an implementation of a distributed WSN application based on a simplified model of light transmittance through the canopy and validate our approach in various field campaigns exemplarily conducted in maize cultivars. Since a ground truth LAI is very difficult to obtain, we use the LAI-2200, one of the most widely established standard instruments, as a reference. We evaluate the accuracy of LAI estimates derived from the analysis of PAR sensor data and the robustness of our sensor modification. As a result, an extensive comparative analysis emphasizes a strong linear correlation ( r 2 = 0.88 , RMSE=0.28) between both approaches. Hence, the proposed WSN-based approach enables a promising alternative for a flexible and continuous LAI monitoring.

On the potential of Wireless Sensor Networks for the in-field assessment of bio-physical crop parameters

The exploration of bio-physical crop parameters is fundamental for the efficiency of smart agriculture. The leaf area index (LAI) is one of the most important crop parameters and serves as a valuable indicator for yield-limiting processes. It contributes to situational awareness ranging from agricultural optimization to global economy. In this paper, we investigate the potential of Wireless Sensor Networks (WSNs) for the in-field assessment of bio-physical crop parameters. Our experiences using commercial off-the-shelf (COTS) sensor nodes for the indirect and nondestructive LAI estimation are described. Furthermore, we present the design of our measurement architecture and results of various in-field measurements. By directly comparing the results achieved by WSN technology with those of a conventional approach, represented by a widely used standard instrument, we analyze whether bio-physical crop characteristics can be derived from WSN data with a desired accuracy. Moreover, we propose a simple approach to significantly enhance the accuracy of COTS sensor nodes for LAI estimation while, at the same time, reveal open challenges.

Improved crop classification using multitemporal RapidEye data

Land Use/Land Cover (LU/LC) of agricultural areas derived from remotely sensed data still remains very challenging. With regard to the rising availability and the improving spatial resolution of satellite data, multitemporal analyses become increasingly important for remote sensing investigations. Even crops with similar spectral behaviour can be separated by adding spectral information of different phenological stages. Hence, the potential of multi-date RapidEye data for classifying numerous agricultural classes was investigated in this study. In an agricultural area in Northern Israel two complete crop cycles 2013 and 2014 with two cultivation periods each were investigated. In order to avoid a high number of classification runs, a pre-procedure was tested to get the multitemporal data set which provides best spectral separability. Therefore, Jeffries-Matusita (JM) measure was used in order to obtain the best multitemporal setting of all available images within one cultivation period. Eight classifiers were applied to compare the potential of separating crops. The three algorithms Maximum Likelihood (ML), Random Forest (RF) and Support Vector Machine (SVM) outperformed by far the other classifiers with Overall Accuracies higher than 90 %. The processing time of ML and RF, however, was significantly shorter compared to SVM, in fact by a factor of five to seven.

Smart fLAIr: A smartphone application for fast LAI retrieval using Ambient Light Sensors

The efficiency of precision agriculture fundamentally depends on the exploration of bio-physical and bio-chemical plant parameters and the assessment of current crop conditions. The leaf area index (LAI) represents one of the most important crop parameters and is defined as the ratio of foliage area to ground area. It is widely-used in agriculture and agronomy as it indicates yield-limiting processes. In this paper, we present Smart fLAIr (fast LAI retrieval), a novel smartphone application for a low-cost in-situ LAI estimation. This estimation is based on the gap fraction analysis, a widespread indirect and non-destructive methodology. For that purpose, Smart fLAIr leverages the smartphones internal Ambient Light Sensor (ALS). However, in order to improve the gap fraction accuracy, we enhance the ALS by a diffuser cap combined with an optical band-pass filter. Our prototype is implemented on the Android platform with a focus on usability aspects and its practicability. Conducting a comparative analyses with a commercial instrument, we successfully evaluated this prototype for maize canopies. The convincing performance of our approach in terms of accuracy and stability highlights the potential of Smart fLAIr as a valuable alternative for in-situ LAI assessment.

Using hyperspectral remote sensing data for the assessment of topsoil organic carbon from agricultural soils

Detecting soil organic carbon (SOC) changes is important for both the estimation of carbon sequestration in soils and the development of soil quality. During a field campaign in May 2011 soil samples were collected from two agricultural fields northwest of Koethen (Saxony-Anhalt, Germany) and the SOC content of the samples was determined in the laboratory afterwards. At the same time image data of the test site was acquired by the hyperspectral airborne scanner AISA-DUAL (450-2500 nm). The image data was corrected for atmospheric and geometric effects and a spectral binning has been performed to improve the signal-to-noise ratio (SNR). For parameter prediction, an empirical model based on partial least squares regression (PLSR) was developed from AISA-DUAL image spectra extracted at the geographic location of the soil samples and analytical laboratory results. The obtained SOC concentrations from the AISA-DUAL data are in accordance with the concentration range of the chemical analysis. For this reason, the PLSR-model has been applied to the AISA-DUAL image data. The predicted SOC concentrations reflect the spatial conditions of the two investigated fields. The results indicate the potential of the used method as a quick screening tool for the spatial assessment of SOC, and therefore an appropriate alternative to time- and cost-intensive chemical analysis in the laboratory.

Towards in-situ sensor network assisted remote sensing of crop parameters: poster

Remote sensing data acquired from satellites are a vital information source for precision agriculture to assess current crop conditions. Field measurements of plant parameters, like the leaf area index (LAI), serve as a crucial basis to validate parameter maps derived from satellite images. Traditionally, in-situ LAI measurements are collected manually. Therefore, the assessment is cost-intensive and the temporal availability of measurements is limited. Measurements provided by small sensor devices organized in a wireless sensor network (WSN) are a low-cost alternative to manual field measurements. They allow a precise LAI determination with high temporal resolution at many different locations in a field or even an entire region. These information are highly demanded for the validation of spatial information on crop conditions derived from image data acquired by modern satellites like Sentinel-2.