Joel O. Paz

Examples of strategies to analyze spatial and temporal yield variability using crop models

Process-oriented crop growth models simulate plant growth over homogeneous areas. The advent of precision farming has resulted in the need to extend the use of point-based crop models to account for spatial processes. Spatial processes include surface and subsurface water flow and spatial and temporal interaction of plant growth with soil water, nutrient and pest stress and management practices. Our research has focused on developing methods to account for spatial interactions in the CROPGRO-Soybean and CERES-Maize models. These methods introduce new challenges for accurately and economically defining spatial inputs for the models. In spite of these challenges, both models have been used to evaluate causes of spatial yield variability with reasonable success. The purpose of this paper is to present several examples of strategies that we have found useful in using these models to assess spatial and temporal yield variability over different environmental conditions and to analyze economic return of prescriptions. Strategies to overcome spatial resolution in point-based crop models include calibration techniques to run point-based models at small scales within a field, using remote sensing to target measurements of models inputs to areas of similar plant response, and linking point-based models to three-dimensional water flow models to better represent water transport. Each strategy is demonstrated using case studies and comparison of simulated and measured data are presented. A method to estimate break-even costs associated with variable soybean cultivar placement in a field is outlined and presented as a case study as well. Crop models can provide useful estimates of potential economic return of prescriptions, as well as estimate the sensitivity of a prescription to weather. They can also estimate the value of weather information on management prescriptions.

ANALYSIS OFWATER STRESS EFFECTS CAUSING SPATIAL YIELD VARIABILITY IN SOYBEANS

Soybean yields have been shown to be highly variable across fields. Past efforts to correlate yield in small sections of fields to soil type, elevation, fertility, and other factors in an attempt to characterize yield variability has had limited success. In this article, we demonstrate how a process oriented crop growth model (CROPGRO-Soybean) can be used to characterize spatial yield variability of soybeans, and to test hypotheses related to causes of yield variability. In this case, the model was used to test the hypothesis that variability in water stress corresponds well with final soybean yield variability within a field. Soil parameters in the model related to rooting depth and hydraulic conductivity were calibrated in each of 224 grids in a 16-ha field in Iowa using three years of yield data. In the best case, water stress explained 69% of the variability in yield for all grids over three years. The root mean square error was 286 kg ha–1 representing approximately 12% of the three-year mean measured yield. Results could further be improved by including factors that were not measured, such as plant population, disease, and accurate computation of surface water run on into grids. Results of this research show that it is important to include measurements of soil moisture holding capacity, and drainage characteristics, as well as root depth as data layers that should be considered in any data collection effort.

Tools for optimizing management of spatially-variable fields

Abstract Efficient use of agro-chemicals is beneficial for farmers as well as for the environment. Spatial and temporal optimization of farm management will increase productivity or reduce the amount of agro-chemicals. This type of management is referred to as Precision Agriculture. Traditional management implicitly considers any field to be a homogeneous unit for management: fertilization, tillage and crop protection measures, for example, are not varied within a single field. The question for management is what to do when . Because of the variability within the field, this implies inefficient use of resources. Precision agriculture defines different management practices to be applied within single, variable fields, potentially reducing costs and limiting adverse environmental side effects. The question is not only what and when but also where . Many tools for management and analysis of spatial variable fields have been developed. In this paper, tools for managing spatial variability are demonstrated in combination with tools to optimize management in environmental and economic terms. The tools are illustrated on five case studies ranging from (1) a low technology approach using participatory mapping to derive fertilizer recommendations for resource-poor farmers in Embu, Kenya, (2) an example of backward modelling to analyze fertilizer applications and restrict nitrogen losses to the groundwater in the Wieringermeer in The Netherlands, (3) a low-tech approach of precision agriculture, developed for a banana plantation in Costa Rica to achieve higher input use efficiency and insight in spatial and temporal variation, (4) a high-tech, forward modelling approach to derive fertilizer recommendations for management units in Zuidland in The Netherlands, and (5) a high-tech, backward modelling approach to detect the relative effects of several stress factors on soybean yield.

ESTIMATING SPATIALLY VARIABLE SOIL PROPERTIES FOR APPLICATION OF CROP MODELS IN PRECISION FARMING

Crop models have been useful for identifying underlying causes of yield variability and evaluating management prescriptions. However, estimating the spatial soil inputs required to calibrate crop models to historic yields has proven to be challenging and time consuming. Currently, calibration techniques require excessive computer time when applied over many grid points within a field, and procedures for estimating unknown inputs are not well defined. The objectives of this research were: (1) to develop an efficient procedure for estimating spatially variable soil properties for the CROPGRO–Soybean model, and (2) to demonstrate its use in diagnosing areas in the field where excess water or water stress reduce soybean yield. A study was conducted for a 12–ha field in Linn County, Iowa, using soybean data collected during two years (1996 and 1998). Yield, soil type, topography, and soil characterization data were used to estimate spatial variations in soil drainage factors (saturated hydraulic conductivity of an impeding layer and tile drainage spacing), water availability (SCS curve number and maximum rooting depth), and a soil fertility factor. A procedure was developed to create a database of predicted yields for combinations of coefficients, and to search the database using rules based on soil classification, drainage class, and topography to guide the parameter estimation process. When rules based on drainage class were used, the CROPGRO–Soybean model explained 45% to 70% of the yield variability for 1996 and 1998, respectively. When rules based on soil water availability, drainage characteristics, and topography were used, good predictions were obtained in both years (r2 = 0.70 for 1996 and 0.80 for 1998), and RMSE was 2.8% of grid level yields. The data base approach required less than half the time that simulated annealing required for the field with 48 grids.

USING CROSS-VALIDATION TO EVALUATE CERES-MAIZE YIELD SIMULATIONS WITHIN A DECISION SUPPORT SYSTEM FOR PRECISION AGRICULTURE

Crop growth models have recently been implemented to study precision agriculture questions within the framework of a decision support system (DSS) that automates simulations across management zones. Model calibration in each zone has occurred by automatically optimizing select model parameters to minimize error between measured and simulated yield over multiple growing seasons. However, to date, there have been no efforts to evaluate model simulations within the DSS. In this work, a model evaluation procedure based on leave-one-out cross-validation was developed to explore several issues associated with the implementation of CERES-Maize within the DSS. Five growing seasons of measured yield data from a central Iowa cornfield were available for cross-validation. Two strategies were used to divide the study area into management zones, one based on soil type and the other based on topography. The decision support system was then used to carry out the model calibration and validation simulations as required to complete the cross-validation procedure. Results demonstrated that the models ability to simulate corn yield improved as more growing seasons were used in the cross-validation. For management zones based on topography, the average root mean squared error of prediction (RMSEP) from cross-validations was 1460 kg ha-1 when two growing seasons were used and 998 kg ha-1 when five years were used. Model performance was shown to vary spatially based on soil type and topography. Average RMSEP was 1651 kg ha-1 on zones of Nicollet loam, while it was 496 kg ha-1 on zones of Canisteo silty clay loam. Spatial patterns also existed between areas of higher RMSEP and areas where measured spatial yield variability was related to topography. Changes in the mean and variance of optimum parameter sets as more growing seasons were used in cross-validation demonstrated that the optimizer was able to arrive at more stable solutions in some zones as compared to others. Results suggested that cross-validation was an appropriate method for addressing several issues associated with the use of crop growth models within a DSS for precision agriculture.

RELATIONSHIP BETWEEN PLANT AVAI LABLE SOIL WATER AND YIELD FOR EXPLAINING SOYBEAN YIELD VARIABILITY

Spatial patterns of crop yield differ from year to year because of spatial and temporal interactions that occur within a field. A clear understanding of spatial soil–water uptake by plant roots is fundamental to understand yield variability and to make management recommendations that maximize profit or minimize environmental impacts. The objective of this study was to investigate variations in water relations within and between soil map units in a field in order to explain spatial distribution of soybean yield. This research was conducted in a 20–ha field in Boone County, Iowa, in 2000. Spatial distribution of soil water was investigated in 30 sites across field using a tube–access TDR probe. Aerial digital photos were taken three times during the growing season to investigate the relationship between plant canopy and resulting yield. Results showed that soybean yield was greatly reduced in the field compared to an average year, probably due to the occurrence of a drier than normal year. The yield variation was about 24%, likely due variation in soil water during pod filling. Soil water balance calculations for selected sites showed that plants likely experienced water stress in mid–July, but the level of stress increased dramatically later in the season and reached its maximum at the end of August. The sites exposed to earlier water stress exhibited lower yield. There was a good correlation (r 2 > 0.48) between plant available soil water and yield for any date during the reproductive phase of the soybean crop. The soil water relations were able to explain more than 48% of yield variability in 30 sites. However, the vegetation index did not correlate well with yield for any of the dates on which remotely sensed images were taken. This poor relationship indicated the variable drought stress that dominated yield variability occurred after full canopy was reached and primarily affected pod numbers, not canopy biomass.

A web-based fuzzy expert system for frost warnings in horticultural crops

Frost damage is responsible for more economic losses than any other weather related phenomenon in the United States (USA) and many other regions across the globe. With sufficient warning, producers can minimize the potential damages caused by frost and freeze events. However, the severity of these events is dependent upon several factors including air temperature, dew point temperature, and wind speed. Methods for assessing this risk are not easily quantifiable and require the insight of experts familiar with the process. Georgias Extreme-weather Neural-network Informed Expert (GENIE) incorporates the knowledge of expert agrometeorologists and additional information on air temperature, dew point temperature, and wind speed into a fuzzy expert system for use by Georgia producers to provide warning levels of frost and freeze for blueberries and peaches. Artificial neural network (ANN) predictions of air temperature and dew point temperature across the state of Georgia for one to 12 h ahead and observed wind speed are used as input variables for this fuzzy expert system. Meteorological conditions were classified into five levels of frost and freeze by the expert agrometeorologists. These expertly classified scenarios were then used to develop fuzzy logic rules and membership functions for GENIE. Additional scenarios were presented to GENIE for evaluation and it classified all scenarios correctly. This tool will be made available to Georgia producers through a web-based interface, which can be found at www.georgiaweather.net.

Analysis of the Inter-Annual Variation of Peanut Yield in Georgia Using a Dynamic Crop Simulation Model

It is common practice to use crop simulation models and long-term weather data to study the impact of climate variability on yield. Simulated yields mainly reflect the weather variability but not the adoption of new technologies; both sources of variation are reflected in long-term observed yields. Therefore, long-term observed yields, if available, cannot be readily used for evaluation of crop models. The objectives of this study were to analyze the impact of climate variability on long-term historical peanut yield in Georgia obtained with a dynamic crop simulation model and to assess the applicability of using long-term average county yield determined from statistical estimates for evaluation of the simulated yield. Observed yields obtained from state variety trials as well as yield estimates from the USDA-NASS for three counties in the Georgia peanut belt from 1934 to 2003 were used for evaluating simulated yield series. Simulated yields based on the CSM-CROPGRO-Peanut model were categorized into three technological periods (TP). A weighted average based on the acreage of the soil type, the peanut type, and the irrigated land in each county was calculated to obtain a unique simulated yield. Then yields and weather data of the 70-year period were grouped with respect to El Nino Southern Oscillation phases and TPs. Pearsons coefficient of correlation, the least significant difference (LSD), and the t-test were used to evaluate the results. When compared with observed yields, NASS estimates failed to estimate the weather variability at the beginning of the period, but simulated yields clearly reflected that variability during the 70-year period. NASS yield estimates seemed to be useful for evaluating simulated yields from the mid-1970s. The results showed that crop models can be useful in understanding the inter-annual variation of yield due to climate variability if appropriate adjustments are made to account for changes and improvements in agrotechnology.

Remote Sensing and Geospatial Technological Applications for Site-specific Management of Fruit and Nut Crops: A Review

Site-specific crop management (SSCM) is one facet of precision agriculture which is helping increase production with minimal input. It has enhanced the cost-benefit scenario in crop production. Even though the SSCM is very widely used in row crop agriculture like corn, wheat, rice, soybean, etc. it has very little application in cash crops like fruit and nut. The main goal of this review paper was to conduct a comprehensive review of advanced technologies, including geospatial technologies, used in site-specific management of fruit and nut crops. The review explores various remote sensing data from different platforms like satellite, LIDAR, aerial, and field imaging. The study analyzes the use of satellite sensors, such as Quickbird, Landsat, SPOT, and IRS imagery as well as hyperspectral narrow-band remote sensing data in study of fruit and nut crops in blueberry, citrus, peach, apple, etc. The study also explores other geospatial technologies such as GPS, GIS spatial modeling, advanced image processing techniques, and information technology for suitability study, orchard delineation, and classification accuracy assessment. The study also provides an example of a geospatial model developed in ArcGIS ModelBuilder to automate the blueberry production suitability analysis. The GIS spatial model is developed using various crop characteristics such as chilling hours, soil permeability, drainage, and pH, and land cover to determine the best sites for growing blueberry in Georgia, U.S. The study also provides a list of spectral reflectance curves developed for some fruit and nut crops, blueberry, crowberry, redblush citrus, orange, prickly pear, and peach. The study also explains these curves in detail to help researchers choose the image platform, sensor, and spectrum wavelength for various fruit and nut crops SSCM.

ESTIMATING POTENTIAL ECONOMIC RETURN FOR VARIABLE SOYBEAN VARIETY MANAGEMENT

This article describes the development and economic analysis of soybean management prescriptions for two 20 ha fields in central Iowa. The Home and McGarvey fields were subdivided into 0.2 ha size grids, and the CROPGRO–Soybean model was calibrated to fit three seasons of historical yield data in each grid. Yields of 70 soybean varieties were estimated using the calibrated model and 34–years of historical weather data. Several variety prescription scenarios were developed and evaluated. Variety prescription based on implementing the best variety for each grid for each year (prescription A) produced higher yields and higher net returns than using the 34–year average–yield prescription (prescription B) or using a uniform maturity group II variety (prescription C) throughout the field. Prescription B is a viable option for farmers to use without the burden of having a priori knowledge of weather information associated with implementing prescription A. The results also support the idea of growing a single top–yielding variety across the whole field, and this type of management prescription produced roughly the same level of profitability as that of prescription B. Implementing a prescription tailored to a specific weather type (e.g., El Nino, La Nina) resulted in very small financial gain compared to prescription B. The farmer should look into implementing prescription B regardless of whether the upcoming year would be a La Nina, an El Nino, or neutral year.