The model code is written in Turbo Pascal (both DOS and Windows versions). An advanced user-friendly interface allows users to easily manipulate input files, verify input parameters for range errors and cross compatibility, create simulations, execute single and batch run simulations, customize utputs, produce text and graphical reports, link to spreadsheet programs, and even select a preferred language for the interface text. Simulations can be customized to invoke only those modules of interest for a particular application (e.g., erosion and nitrogen simulation can be disabled if not desired), producing more efficient runs and simplifying model parameterization. The model is fully documented (Stockle and Nelson, 1994, Stockle and Nelson, 1996), and the manual is also available as a help utility from the CropSyst interface. CropSyst executable program, manual, and tutorials can be retrieved directly over the Internet.

*

Because of the variable composition of pig slurry, treatments resulted different during the two years. Nitrogen from pig slurry was subdivided into NH₄ and organic N. Mineral fertilization was applied as urea and then is reported as NH₄. The version 1.08.09 of CropSyst was used to simulate the treatments:

Initial soil water and nitrate content were measured and entered as model inputs.

Results
The results are reported as graphs for soil PAW (plant available water) and soil nitrate (NO₃-N). Data were integrated over depth to represent the layers 0-0.4 m depth and 0.4-0.8 m.

Figure 1. Predicted (solid lines; CS) and measured (symbols; act) PAW data, years 1993-94.










Figure 2. Soil nitrate content during the years 1993-94 for the four treatments simulated. CS (solid lines) gives the predicted values and act (symbols) gives the measured values.

Conclusions
Although CropSyst was able to simulate correctly the nitrogen stress effect on yield and biomass production (data not presented here), some uncertainties showed up in soil NO₃-N simulations when N was applied as urea (1993). Nonetheless, these preliminary results, which require further analysis to improve soil N parameters calibration, show that CropSyst can be a valuable tool to simulate N balance under different management options.

Acknowledgments
PANDA project, Subproject 2, series 2, paper no. 41

*
EVALUATION OF CROPSYST FOR CROPPING SYSTEMS AT TWO LOCATIONS OF NORTHERN AND SOUTHERN ITALY (1996 - study carried out in 1992) paper
We tested the capability of CropSyst, a cropping system simulation model, to simulate cropping systems at two locations (Modena and Foggia) representative of the largest plain areas of Italy. Experimental data collected from rotation experiments during the period 1985-91 were used to calibrate and validate the model. The data set available was not detailed enough to test most of the sub-models, but it was adequate for a satisfactory test of overall model performance. Simulations were conducted for six-year rotations providing initial conditions of the first year without reinitializing the model in later years in the rotation sequence. Model evaluation was performed with the understanding that a large amount of variability was present in the data and that rotation effects other than water (e.g., diseases) were not accounted for by the model. The phenology sub-model proved sufficiently accurate, showing some limitation only in the case of winter cereals. Model simulations of maize, soybean, and barley growth at Modena, and sorghum and sunflower growth at Foggia were reasonably accurate. CropSyst was able to predict neither the yields of soybean grown as a second crop, nor the durum wheat yield variability in the rotations at Foggia. Although CropSyst was able to simulate reasonably well a number of cropping systems, the development of a more detailed field data base is desirable for further evaluation and improvement of the model for Italian conditions.

*

The RadEst program evaluates daily global solar radiation values for a location at a given latitude and estimates daily values. Four models are available to estimate daily radiation from air temperature data; they include and are all derived from the model proposed by Bristow and Campbell (1984). All the models estimate atmospheric transmissivity of global solar radiation based on the difference between maximum and minimum temperature. The estimated value of radiation is calculated as the product of the estimated transmissivity by the value of potential radiation outside the earth atmosphere. Utilities allow a graphical evaluation of the estimates and compute statistical indices for model/location comparison. For each model, one parameter can be fitted using an iterative procedure. Estimated values can be saved as ASCII files.

*

EVALUATION OF METHODS TO ESTIMATE SOIL WATER CONTENT AT FIELD CAPACITY AND WILTING POINT (1996)

M. Donatelli¹, M. Acutis¹, N. Laruccia<sup>3

1</sup> ISA, Sez. di Modena, Modena, Italy
² DASGT, Torino, Italy
³ Cooperativa I. ter, Bologna, Italy

Introduction
Simulation models are effective tools to evaluate the impact of agricultural management; however, their use is often limited because they require a large number of input parameters required to model the mechanistic processes which influence the plant-soil system. In fact, while the input parameter values are generally known in the conditions of research stations, when simulation models are run for large land areas, some input parameters must often be estimated. Two soil parameters which are needed by most simulation models are volumetric water content (SWC) at field capacity (FC) and wilting point (WP). In the literature, several methods to estimate these parameters have been proposed. Most of them have a strong empirical basis, consequently their applicability may be limited to the data set used to define the method. Moreover, comparisons among different methods have seldom been made.

Methods
We developed a program for Windows, SOILPAR, which implements several methods commonly used to estimate soil parameters. The methods were taken from the utility ASW/EPIC and the program LEACHW source code. Data describing 38 soil profiles were used in this study (Table 1).

Table 1. Range of variability for parameters in the 38 soil profiles.

parameter	range (measured values)
water content at wilting point (m3 m-3)	0.020-0.335
water content at field capacity (m3 m-3)	0.109-0.506
sand (%)	2-93
silt (%)	1-52
clay (%)	3-77
organic carbon (%)	0.09-5.83

A preliminary version of the SOILPAR program is available upon request. (see SOILPAR homepage)

Results
The best methods show a low RMSE, both R² and slope close to 1, a high EF, and a CRM close to 0.

Table 2. Key: n = no. of observations; RMSE = root mean squared error^†; EF = modelling efficiency; CRM = coefficient of residual mass; R² = R squared; slope = slope of the line estimated vs. measured.

method	n	RMSE	EF	CRM	R2	slope
Baumer ASW/EPIC	286	0.07	0.62	0.12	0.71	0.83
Brakensiek Rawls LEACHW	286	0.06	0.69	0.01	0.74	0.94
British SS - subsoils LEACHW	286	0.06	0.75	-0.01	0.76	0.85
British SS - topsoils LEACHW	286	0.06	0.70	-0.07	0.75	0.87
EPIC ASW/EPIC	286	0.08	0.57	0.14	0.70	0.80
Hutson LEACHW	286	0.08	0.55	-0.10	0.62	0.51
Manrique ASW/EPIC	286	0.08	0.47	0.20	0.70	0.87
Rawls ASW/EPIC	286	0.06	0.70	0.11	0.77	0.77

Conclusions
The Brakensiek and Rawls, and the British Soil Survey methods, appear to be the most reliable options to estimate SWC at FC and WP. To allow broadening comparison of methods, the implementation of other methods in the program SOILPAR continues.

Acknowledgments
We gratefully acknowledge Dr. M. Guermandi, Servizio Cartografico Emilia Romagna, Bologna, Dr. R. Francaviglia and Dr. D. Bartolini, ISNP, Rome, for providing most of the data used in this study.

---

top