Multiple Cost Functions and Regularization Options for Improved Retrieval of Leaf Chlorophyll Content and LAI through Inversion of the PROSAIL Model

Rivera, Juan Pablo; Verrelst, Jochem; Leonenko, Ganna; Moreno, José

doi:10.3390/rs5073280

Cited by 126 publications

(106 citation statements)

References 62 publications

Supporting

Mentioning

103

Contrasting

Order By: Relevance

“…All processing modules are accessible by drop-down menus from the top bar. Already in the first version of ARTMO, LUT-based inversion applications were available [12,13,40]. ARTMO has been extended since then and v3 is formally presented in this paper.…”

Section: Artmo Software Packagementioning

confidence: 99%

“…They are based on parametric and non-parametric regression as well as physically-based inversion using a LUT. This has led to the development of a: (1) "Spectral Indices assessment toolbox"; (2) "Machine Learning Regression Algorithm toolbox" [41]; and (3) "LUT-based inversion toolbox" [12]. Figure 1.…”

Section: Artmo Software Packagementioning

confidence: 99%

“…A statistical regression model enables the linkage of EO data with biophysical parameters of interest [10] by making use of (in situ) calibration data. On the other hand, physically-based approaches allow the retrieval of biophysical parameters by inversion of a radiative transfer model (RTM) against EO data [11,12].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

On the Semi-Automatic Retrieval of Biophysical Parameters Based on Spectral Index Optimization

et al. 2014

Self Cite

View full text Add to dashboard Cite

Regression models based on spectral indices are typically empirical formulae enabling the mapping of biophysical parameters derived from Earth Observation (EO) data. Due to its empirical nature, it remains nevertheless uncertain to what extent a selected regression model is the most appropriate one, until all band combinations and curve fitting functions are assessed. This paper describes the application of a Spectral Index (SI) assessment toolbox in the Automated Radiative Transfer Models Operator (ARTMO) package. ARTMO enables semi-automatic retrieval and mapping of biophysical parameters from optical remote sensing observations. The SI toolbox facilitates the assessment of biophysical parameter retrieval accuracy of established as well as new and generic SIs. For instance, based on the SI formulation used, all possible band combinations of formulations with up to ten bands can be defined and evaluated. Several options are available in the SI assessment: calibration/validation data partitioning, the addition of noise and the definition of curve fitting models. To illustrate its functioning, all two-band combinations according to simple ratio (SR) and normalized difference (ND) formulations as well as various fitting functions (linear, exponential, power, logarithmic, polynomial) functions led to similar best correlations, though spatial differences emerged when applying the functions to HyMap imagery. We suggest that a systematic SI assessment is a strong requirement in the quality assurance approach for accurate biophysical parameter retrieval.

show abstract

Section: Artmo Software Packagementioning

confidence: 99%

Section: Artmo Software Packagementioning

confidence: 99%

See 1 more Smart Citation

On the Semi-Automatic Retrieval of Biophysical Parameters Based on Spectral Index Optimization

et al. 2014

Self Cite

View full text Add to dashboard Cite

show abstract

“…reflectance and transmittance) parameterized by the following inputs: chlorophyll content (C ab ) in µg/cm 2 , leaf dry mass per unit area (C m ) in mg/cm 2 , leaf water mass per unit area (C w ) in mg/cm 2 , and effective number of leaf layers (N) (Jacquemoud and Baret 1990). The model has been widely applied to broadleaf vegetation to estimate chlorophyll content (Zhang et al 2008, Ma et al 2012, Rivera et al 2013. It has also been successfully recalibrated and used to simulate the optical properties of coniferous needles (Malenovský et al 2008, Morsdorf et al 2009).…”

Section: Prospect Model Simulationmentioning

confidence: 99%

Prospect inversion for indirect estimation of leaf dry matter content and specific leaf area

Ali

Darvishzadeh

Skidmore

et al. 2015

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:Quantification of vegetation properties plays an indispensable role in assessments of ecosystem function with leaf dry mater content (LDMC) and specific leaf area (SLA) being two important vegetation properties. Methods for fast, reliable and accurate measurement of LDMC and SLA are still lacking. In this study, the inversion of the PROSPECT radiative transfer model was used to estimate these two leaf parameters. Inversion of PROSPECT traditionally aims at quantifying its direct input parameters rather than identifying the parameters which can be derived indirectly from the input parameters. The technique has been tested here to indirectly model these parameters for the first time. Biophysical parameters such as leaf area, as well as fresh and dry weights of 137 leaf samples were measured during a field campaign in July 2013 in the mixed mountain forests of the Bavarian Forest National Park, Germany. Reflectance and transmittance of the leaf samples were measured using an ASD field spec III equipped with an integrating sphere. The PROSPECT model was inverted using a look-up table (LUT) approach for the NIR/SWIR region of the spectrum. The retrieved parameters were evaluated using their calculated R 2 and normalized root mean square error (nRMSE) values with the field measurements. Among the retrieved variables the lowest nRMSE (0.0899) was observed for LDMC. For both traits higher R 2 values (0.83 for LDMC and 0.89 for SLA) were discovered. The results indicate that the leaf traits studied can be quantified as accurately as the direct input parameters of PROSPECT. The strong correlation between the estimated traits and the NIR/SWIR region of the electromagnetic spectrum suggests that these leaf traits could be assessed at canopy and in the landscape by using hyperspectral remote sensing data.

show abstract

“…The toolbox has been improved and expanded since then, such as by the implementation of retrieval modules. They are based on parametric and non-parametric regression, as well as physically-based inversion using a LUT and led to the development of a: (1) "Spectral Indices assessment toolbox" [37]; (2) "Machine Learning Regression Algorithm (MLRA) toolbox" [57]; and (3) "LUT-based inversion toolbox" [58]. ARTMO v3.10 with the "Emulator toolbox v. 1.00" is formally presented in this paper.…”

Section: Artmomentioning

confidence: 99%

An Emulator Toolbox to Approximate Radiative Transfer Models with Statistical Learning

et al. 2015

Self Cite

View full text Add to dashboard Cite

Physically-based radiative transfer models (RTMs) help in understanding the processes occurring on the Earth's surface and their interactions with vegetation and atmosphere. When it comes to studying vegetation properties, RTMs allows us to study light interception by plant canopies and are used in the retrieval of biophysical variables through model inversion. However, advanced RTMs can take a long computational time, which makes them unfeasible in many real applications. To overcome this problem, it has been proposed to substitute RTMs through so-called emulators. Emulators are statistical models that approximate the functioning of RTMs. Emulators are advantageous in real practice because of the computational efficiency and excellent accuracy and flexibility for extrapolation. We hereby present an "Emulator toolbox" that enables analysing multi-output machine learning regression algorithms (MO-MLRAs) on their ability to approximate an RTM. The toolbox is included in the free-access ARTMO's MATLAB suite for parameter retrieval and model inversion and currently contains both linear and non-linear MO-MLRAs, namely partial least squares regression (PLSR), kernel ridge regression (KRR) and neural networks (NN). These MO-MLRAs have been evaluated on their precision and speed to approximate the soil vegetation atmosphere transfer model SCOPE (Soil Canopy Observation, Photochemistry and Energy balance). SCOPE generates, amongst others, Remote Sens. 2015, 7 9348 sun-induced chlorophyll fluorescence as the output signal. KRR and NN were evaluated as capable of reconstructing fluorescence spectra with great precision. Relative errors fell below 0.5% when trained with 500 or more samples using cross-validation and principal component analysis to alleviate the underdetermination problem. Moreover, NN reconstructed fluorescence spectra about 50-times faster and KRR about 800-times faster than SCOPE. The Emulator toolbox is foreseen to open new opportunities in the use of advanced RTMs, in which both consistent physical assumptions and data-driven machine learning algorithms live together.

show abstract

Multiple Cost Functions and Regularization Options for Improved Retrieval of Leaf Chlorophyll Content and LAI through Inversion of the PROSAIL Model

Cited by 126 publications

References 62 publications

On the Semi-Automatic Retrieval of Biophysical Parameters Based on Spectral Index Optimization

On the Semi-Automatic Retrieval of Biophysical Parameters Based on Spectral Index Optimization

Prospect inversion for indirect estimation of leaf dry matter content and specific leaf area

An Emulator Toolbox to Approximate Radiative Transfer Models with Statistical Learning

Contact Info

Product

Resources

About