User Friendly Corn Yield Prediction Model

Search PMN

E. C. A. Runge, Professor and Billie Turner Chair in Production Agronomy (Emeritus), Soil & Crop Sciences Department, Texas A&M University, College Station 77843-2474; and John F. Benci, Consultant, 382 Moore Road, West St. Paul, Manitoba, Canada R4A 7A2

Abstract

The water balance corn yield model was made user friendly by adapting it to an Excel spreadsheet. Instead of more direct evapo-transpiration measures, the model utilizes readily available input proxies of maximum daily temperature, rainfall and plant-available stored soil moisture at planting. Maximum daily temperature averaged for each week and total weekly precipitation are used for 10 weeks with 6 weeks before and 4 weeks after tasseling (anthesis). Plant available stored soil moisture at planting is a soil quality factor. High-yielding Midwestern Corn-Belt soils store 10 inches or more of plant-available stored soil moisture at planting. The model has been successfully used to forecast corn yields for individual fields as well as for entire states of Illinois, Iowa, Indiana, Kansas, Michigan, Minnesota, Missouri, Nebraska, Ohio, South Dakota, and Wisconsin for 2001-2005. Model forecasted yields are updated weekly during the growing season as each week of the previous year’s weather is replaced with actual weather. Final forecasted yields are available 4 weeks after tasseling. The model compares the current year’s weather and yield forecast to the previous year’s actual weather and yield. Individual farmers, advisers and scientists will find the model useful in forecasting yields in their own fields.

Introduction

The model utilizes readily available yield and weather data for a previous year and compares it to weather occurring for the current year to forecast a yield (5,6,8,10). Weather data are needed (average weekly maximum air temperature and total weekly precipitation) for 6 weeks before to 4 weeks after tasseling. The model can be used anywhere such data are available. It is much easier to use than other forecasting models (1,2,3). Irrigation can be accommodated by adding water supplied by irrigation to weekly precipitation (10). Plant-available stored soil moisture at planting is a soil quality factor and must be entered to run the model. It can be determined by knowing the soil types for areas where winter rains consistently recharge the profile to field capacity (8,9,10). For drier areas of the western Corn Belt the amount of water in the soil at or near planting time needs to be determined. Rainfall was accumulated from 1 November to planting to help determine plant-available stored soil moisture at planting in drier areas (10). High yielding Corn-Belt soils store 10 inches or more of plant-available stored soil moisture at planting (5).

Most non-irrigated fields used for growing corn have soils that store between 6 and 12 inches of water at planting. However, many fields often include small acreages of soils that store less water (Fig. 1). Farmers are familiar with these lower yielding areas in their fields. Such soils in southern Illinois are often referred to as "slick spots." Other examples of soils storing less plant available stored soil moisture at planting are soils shallow to high bulk density glacial till or bedrock, and sandy soils. Our results indicate that soils with lesser amounts of plant-available stored soil moisture have better and more consistent yields in the northern Corn Belt than they do in the southern Corn Belt (10,11). Within field yield variation is primarily due to differences in plant-available stored soil moisture at planting and yield forecasts for such areas can be made by changing the amount of soil moisture (5,7).

Fig. 1. Effect of variation in plant-available stored soil moisture on corn growth and yield. Corn in background has about 10 inches of PASSM, in left foreground about 4 inches of PASSM, and intermediate amounts in right foreground.

To use the model it is necessary to forecast the tasseling date a month or so after planting. Corn develops more rapidly when temperatures are warm than when it is cool. Under normal conditions tasseling occurs some 10 to 11 weeks after planting. Rainfall and maximum daily temperatures are entered into the model some 4 to 5 weeks after planting. In some years weather conditions may require an adjustment of the forecasted tassel date (10). Water from irrigation can be added to weekly precipitation to determine it’s effect on forecasted corn yield.

Derivation of Model

The model was developed from plot data collected from 1969-1971 from four locations in Illinois (5,6). The hybrids used were DeKalb’s XL 45 and XL 45a, two of the first high population single cross hybrids grown over a wide geographic area. Plant population in the study was 21,425 plants per acre and ranged from 18,225 to 25,717 (5,6). The average plant population for Illinois, Iowa, and Indiana in 1969-1971 was 17,800 while it was 27,400 plants per acre for 2004-2006 (Ty Kalaus, Field Crop Section, NASS-USDA, personal communication). Increasing plant population is one of the main reasons corn yields have been increasing (4). Model results are technologically based on the previous year’s weather, plant-available stored soil moisture at planting, and yield data entered into the model (Fig. 5). The model assumes that corn responds to moisture stress today the same as it did when the model was developed (1969-1971). This may not be valid since many of today’s hybrids may be more drought tolerant than were DeKalb’s XL 45 and XL 45a. Results for 2005 (10) indicate that today’ hybrids are more drought tolerant than were DeKalb’s XL 45 and 45a. If today’s corn hybrids respond differently to weather and moisture stress conditions than did DeKalb’s 45 and 45a the result would not be incorporated in the forecasted yield. If corn growing conditions for the current year are more favorable than they were for the previous year (historical year), forecasted yields will exceed yields for the previous year and vice versa.

The model is on an excel spreadsheet and is easy to use. Spreadsheets 1, 2, and 3 provide the information needed to run the model and how to input data. Ten weeks of average maximum daily temperatures and precipitation for each week are entered on the spreadsheet for the previous (historical) year and also for the current year (Spreadsheet 3). Initially the current and historical year’s rainfall and temperature data are the same on the spreadsheet (Spreadsheet 3). As the growing season progresses and weather data become available, the previous year’s weather data is replaced by the current year’s data (Spreadsheet 3). Differences between weather data for the current and previous year will change the current year’s forecasted corn yield. More favorable weather in the current than in the previous year will give higher forecasted yields and vice versa. The model has been successfully used to forecast corn yields (10). Models lends themselves to answering "what if" questions. For example, what impact would 95°F temperatures coupled with no rain have on forecasted yield? Or what impact would a 2-inch rain or irrigation during the tasseling week have on forecasted corn yield?

Sensitivity Analysis

Before the model was used extensively, it was tested in three counties for each of the 11 states (Illinois, Iowa, Indiana, Kansas, Michigan, Minnesota, Missouri, Nebraska, Ohio, South Dakota, and Wisconsin) for the period from 1960-1998 (Fig. 2). The test determined if model forecasted yields mimicked the effect of past weather on reported corn yields (www.nass.usda.gov) for the three counties in each state. The county comparisons should also identify any problems that might occur when using the model. The analysis compared actual to model calculated corn yields for 33 counties (10). Within a county, model calculated corn yields were made using the same planting date and the same plant available stored soil moisture at planting for all years. Planting date data were not available to calculate tassel dates for individual years. Only one weather station was used to represent each county’s weather.

Fig. 2. County actual (yact) and simulated model (ymodel) corn yields for 1960-1998 for selected counties: Champaign Co., IL; New Madrid Co., MO; Union Co., SD; and Jackson Co., WI.

Spreadsheet 1. Read me first screen (Excel file).

Spreadsheet 2. Plant-available stored soil moisture at planting screen (Excel file).

Spreadsheet 3. Data input screen (PowerPoint file).

Figure 2 gives actual and model simulated yields for Champaign Co., IL; New Madrid Co., MO; Union Co., SD; and Jackson Co., WI. Model simulated yields were regressed using the trend line for actual county yields. Results were encouraging and model yields mimicked NASS-USDA reported county yields for the three counties in all eleven states of the Corn Belt for the 1960-1998 period. The largest R², 92%, between model and actual county yields occurred for New Madrid Co., MO (Fig. 2).

Steps in Using the Excel Spreadsheet Model/Calculator

The model on the excel spreadsheet is available for download free of charge by request to the corresponding author, E. C. A. Runge. It can be freely distributed without further consultation. To familiarize readers with the model three excel spreadsheet screens are presented in Spreadsheets 1, 2, and 3.

(i) Read Me Screen (Spreadsheet 1). The Read Me First screen gives information on:

• Saving the spreadsheet on your computer;

• Requirements to run the forecasting model;

• How to begin;

• Results; and

• Assumptions made in using the corn yield calculator or model.

(ii) PASSM Screen (Spreadsheet 2). The plant available stored soil moisture at planting screen gives information on:

• Guidelines for selecting plant-available stored soil moisture at planting for your field or parts of your field; and

• Table or index for plant-available stored soil moisture at planting for Corn-Belt soils.

(iii) Data Input Screen (Spreadsheet 3). The date input screen gives information on entering data needed to run the model and calculates the current year’s forecasted yield. It also gives the percent the current year’s yield forecast is compared to the previous year’s yield.

• Data entry and yield forecast page for the model;

• Previous year’s weather, plant-available stored soil moisture at planting and corn yield;

• This year’s weather, plant-available store soil moisture at planting, forcasted yield for current year and percent the current year yield is versus the previous year’s yield.

Conclusion

The model has been tested extensively in the 11 Corn-Belt states and for the transect from Raymondville, TX, to Brookings, SD. Results mimics actual corn yields reported by NASS-USDA (10). The model is available free of charge to anyone desiring a copy upon request to the corresponding author, E. C. A. Runge.

Literature Cited

1. Anapalli, S. S., Ma, L., Nielsen, D. C., Vigil, M. F., and Ahuja, L. R. 2005. Simulating planting date effects on corn production using RZWQM and DERES-maize models. Agron. J. 97:58-71.

2. Baez-Gonzalez, A. D., Kiniry, J. R., Maas, S. J., Tiscareno, M. L., Macias, C. J., Mendoza, J. L., Richardson, C. W., Salinas, G. J., and Manjarrez, J. R. 2005. Large-area maize yield forecasting using leaf area index based yield model. Agron. J. 97:418-425.

3. Gerik, Tom. 2008. Crop weather analyzer. Online. CroPMan: Crop Production and Management Model. Blackland Res. and Ext. Center, Temple, TX.

4. Duvick, D. N., and Cassman, K. G. 1999. Post-green revolution trends in yield potential of temperate maize in the north-central United Statees. Crop Sci. 39:1622-1630.

5. Leeper, R. A., Runge, E. C. A., and Walker, W. M. 1974. Effect of plant-available stored soil moisture on corn yields. I. Constant climatic conditions. Agron. J. 66:723-728.

6. Leeper, R. A., Runge, E. C. A., and Walker, W. M. 1974. Effect of plant-available stored soil moisture on corn yields. II. Variable climatic conditions. Agron. J. 66:728-733.

7. Morgan, C. L. S., Norman, J. M., and Lowery, B. 2003. Estimating plant-available water across a field with and inverse yield model. Soil Sci. Soc. Am. J. 67:620-629.

8. Runge, E. C. A., and Benci, J. F. 1975. Modeling corn production: Estimating production under variable soil and climatic conditions. Pages 194-214 in: Proc. of the 30th Ann. Corn and Sorghum Res. Conf. Am. Seed Trade Assoc. Washington, DC.

9. Runge, E. C. A., and Benci, J. F. 2008. Climate change: Has climate become more or less favorable for growing corn over the past century?. Online. Crop Management doi:10.1094/CM-2008-0609-02-RS.

10. Runge, E. C. A., and Benci, J. F. 2008. Forecasting corn yield for eleven states in the corn belt: Results for 2001-2005. Online. Crop Management doi:10.1094/CM-20080609-03-RS.

11. Runge, E. C. A., and Benci, J. F. 2008. Climate change: Effect of 1922-2002 weather on corn yield for a transect from Raymondville, TX, to Brookings, SD. Online. Crop Management doi:10.1094/CM-2008-0609-01-RS.