Martine Guérif

Remote sensing and crop production models: present trends

Abstract The use of remote sensing as a tool for estimating crop production on a regional scale has been suggested for quite a while. Since the use of remotely sensed information to directly estimate crop production is questionable (primarily due to the indirect link between remotely sensed data and crop state variables), crop models may be used as a companion tool to remote sensing. Various types of crop models (statistical, deterministic, semi-empirical) are described, and specific methods (forcing, recalibration, statistical correction) are described for introducing remotely sensed information into models. The appropriance of combining different model types and sources of remotely sensed information, problems related to time scales, and the need for robust and independent phenological routines are also discussed.

Global sensitivity analysis measures the quality of parameter estimation: The case of soil parameters and a crop model

One common limitation of the use of crop models for decision making in precise crop management is the need for accurate values of soil parameters for a whole field. Estimating these parameters from data observed on the crop, using a crop model, is an interesting possibility. Nevertheless, the quality of the estimation depends on the sensitivity of model output variables to the parameters. The goal of this study is to explain the results for the quality of parameter estimation based on global sensitivity analysis (GSA). The case study consists of estimating the soil parameters by using the STICS-wheat crop model and various synthetic observations on wheat crops (LAI, absorbed nitrogen and grain yield). Suitable criteria summarizing the sensitivity indices of the observed variables were created in order to link GSA indices with the quality of parameter estimation. We illustrate this link on 16 different configurations of different soil, climatic and crop conditions. The GSA indices were computed by the Extended FAST method and a function of RMSE was computed with an importance sampling method based on Bayes theory (GLUE). The proposed GSA-based criteria are able to rank the parameters with respect to their quality of estimation and the different configurations (especially climate and observation set) with respect to their ability to estimate the whole parameter set. They may be used as a tool for predicting the performance of different observation datasets with regard to parameter estimation.

Calibration of the SUCROS emergence and early growth module for sugar beet using optical remote sensing data assimilation

Crop models are useful for monitoring crop production on a local scale. Their application to a larger area, such as a region, is hampered by the difficulty in determining the value of some of their parameters, which may differ greatly between fields. The use of optical remote sensing helps to overcome this problem. Coupling a radiation transfer model to a crop model makes it possible to simulate reflectance for those times in crop growth for which remote sensing data are available. The inversion of the combined model on these data then makes it possible to estimate new values for certain sensitive parameters of the crop model. This paper describes the use of such a method on a local scale, for sugar beet, focusing on the parameters describing emergence and early crop growth. These processes vary greatly depending on the soil, climate and seedbed preparation, and affect yield significantly. The SUCROS crop model and the SAIL reflectance model were combined. The resulting model was calibrated under standard conditions and then evaluated under test conditions to which the emergence and early growth parameters of the SUCROS model were adjusted. The test conditions seedbed structure was coarser, and the sowing depth was greater than expected. Consequently, emergence occurred later, and the initial leaf area was smaller. The SUCROS simulation using standard values for emergence and early growth parameters did not accurately predict crop growth under these test conditions. The inversion of the combined model using a set of canopy reflectance measurements during crop establishment provided new parameter values that allowed us to accurately estimate crop yield. Application of this method on a regional scale, for yield prediction or agronomic diagnosis, should be of great value.

Surface temperature and evapotranspiration: Application of local scale methods to regional scales using satellite data

Abstract Remotely sensed surface temperatures have proven useful for monitoring evapotranspiration (ET) rates and crop water use because of their direct relationship with sensible and latent energy exchange processes. Procedures for using the thermal infrared (IR) obtained with hand-held radiometers deployed at ground level are now well established and even routine for many agricultural research and management purposes. The availability of IR from meteorological satellites at scales from 1 km (NOAA-AVHRR) to 5 km (METEOSAT) permits extension of local, ground-based approaches to larger scale crop monitoring programs. Regional observations of surface minus air temperature (i.e., the stress degree day) and remote estimates of daily ET were derived from satellite data over sites in France, the Sahel, and North Africa and summarized here. Results confirm that similar approaches can be applied at local and regional scales despite differences in pixel size and heterogeneity. This article analyzes methods for obtaining these data and outlines the potential utility of satellite data for operational use at the regional scale.

The use of remotely sensed data in estimation of PAR use efficiency and biomass production of flooded rice

Abstract A model of biomass production for flooded rice crops is proposed based on an energetic yield approach, in which PAR interception efficiency is estimated from vegetation index calculated from canopy reflectances at visible and near-infrared wavelengths. The relationship between interception efficiency (measured from hemispherical photographs) and the vegetation index results from coupling reflectance and PAR interception models based on the Beer-Lambert extinction law. This relationship does not depend explicitly on the canopy structure parameters but it assumes that the background reflectance is known. Although the background reflectance depends on the water state and depth variation in a way which cannot be explicitly quantified, it has been empirically described by its observed evolution fitted with time. After anthesis, this reflectance takes into account the reflectance of the senescent vegetation. Given the spectral profile of a crop (temporal evolution of its reflectance), its PAR absorption profile can be estimated. The PAR use efficiency (conversion of the absorbed energy into biomass) is then determined for every phenological period as the ratio of total absorbed PAR to total synthetized biomass during the period. For flooded rice, root extraction and therefore root biomass evaluation are rather easy, so that more realistic PAR use efficiencies can be estimated than for most crops. The obtained values for these efficiencies agree with those of the literature. Observed varietal differences agree with independent experimental evaluation of cultivar photosynthetic capabilities.

Spectral estimates of the absorbed photosynthetically active radiation and light-use efficiency of a winter wheat crop subjected to nitrogen and water deficiencies†

Abstract A linear regression equation is found relating the photosynthetically active radiation intercepted by the canopy (PARi), measured with hemispherical photographs, and both the normalized difference ND and the ratio NIR/R vegetation indices. On the basis of this equation, NIR/R is used to estimate PARi during the crop cycle. The efficiency with which the PAR absorbed by the crop is transformed into biomass (ϵc) is calculated for three phenological phases of the crop. Nitrogen fertilization is the main factor affecting light interception. At the booting stage, PARi is about 15 per cent greater for treatments with higher nitrogen levels. ϵc is influenced by both nitrogen and irrigation levels, and varies with the phenological phases of the crop. For the irrigated plots, ϵc is higher in the period going from anthesis to soft dough and not in the period from stem elongation to anthesis as most published results indicate. Water stress is the main factor affecting ϵc. The greatest reductions of ;ϵc are o...

A package of parameter estimation methods and implementation for the STICS crop-soil model

Parameter estimation for complex process models used in agronomy or the environmental sciences is important, because it is a major determinant of model predictive power, and difficult, because the models and associated data are complex. Statistics provides guidance for parameter estimation under various assumptions concerning model error, but it is hard to know which assumptions are most acceptable for these models. We therefore propose a collection of parameter estimation methods. All are based on weighted least squares, but different assumptions lead to different weights. The methods allow one to fit simultaneously several different response variables. One can assume that all errors are independent or on the contrary are correlated. One can assume that model error has expectation zero or not. A software package called OptimiSTICS has been developed, that allows one to implement all of the proposed methods with the STICS crop-soil model. The software can in addition treat the case where some parameters are genotype specific while others are common to all genotypes. The software can also automatically do several sequential stages of parameter estimation. An example is presented, which shows the information that can be obtained, and the conclusions drawn, from comparing the different estimation methods.

Crop reflectance estimate errors from the SAIL model due to spatial and temporal variability of canopy and soil characteristics

Abstract Radiative transfer models must be combined with crop growth models for the latter to be recalibrated on each point of a regional domain using remote sensing data. However, radiative transfer models have parameters depending on vegetation and soil characteristics which vary spatially and temporally at a regional scale, but are generally unknown. Their estimation leads to reflectance simulation errors. These errors are analyzed in this article for the visible and near-infrared part of the spectrum using the SAIL model for sugar beets. Two ways for estimating SAIL parameters were compared: One does not require any knowledge of magnitude and variability. The other one is based on rules developed from parameter distribution analysis and extra knowledge on soils and climate. The reflectance estimation error was very high for LAI

Simple and robust methods for remote sensing of canopy chlorophyll content : a comparative analysis of hyperspectral data for different types of vegetation

Canopy chlorophyll content (CCC) is an essential ecophysiological variable for photosynthetic functioning. Remote sensing of CCC is vital for a wide range of ecological and agricultural applications. The objectives of this study were to explore simple and robust algorithms for spectral assessment of CCC. Hyperspectral datasets for six vegetation types (rice, wheat, corn, soybean, sugar beet and natural grass) acquired in four locations (Japan, France, Italy and USA) were analysed. To explore the best predictive model, spectral index approaches using the entire wavebands and multivariable regression approaches were employed. The comprehensive analysis elucidated the accuracy, linearity, sensitivity and applicability of various spectral models. Multivariable regression models using many wavebands proved inferior in applicability to different datasets. A simple model using the ratio spectral index (RSI; R815, R704) with the reflectance at 815 and 704 nm showed the highest accuracy and applicability. Simulation analysis using a physically based reflectance model suggested the biophysical soundness of the results. The model would work as a robust algorithm for canopy-chlorophyll-metre and/or remote sensing of CCC in ecosystem and regional scales. The predictive-ability maps using hyperspectral data allow not only evaluation of the relative significance of wavebands in various sensors but also selection of the optimal wavelengths and effective bandwidths.

Characterizing the spatial and temporal variability of biophysical variables of a wheat crop using hyper-spectral measurements

The spatial extension to field scale of biophysical variables retrieval from radiometric measurements is addressed. Hyper-spectral measurements were acquired over a wheat field on 4 dates in 2000 with a CASI sensor. In 2002, the XYBION multi-spectral sensor (6 bands) was used to acquire images on 5 dates. For each year, ground measurements were performed at ground level to estimate green leaf area index and Cab (leaf chlorophyll content). Two methods of retrieving biophysical variables were tested. The inversion through radiative transfer modeling was used with hyperspectral measurements, whereas empirical relationships were used to estimate variables from XYBION data. The results were discussed in terms of spatial and temporal variations of the estimates.