Crop model used to determine the relationship between climate and crop growth. Crop models that are capable of simulating the effects of weather, soil type, seed variety, fertilizer management and irrigation on crop growth and development. Results from such simulations can then be used to predict changes and detect trends in biophysical indicators such as crop yield, nutrient uptake, nitrate leaching and soil carbon levels. Once set within the framework of a comprehensive information and decision support system, the crop models can facilitate the effective analysis of issues related to agricultural production, resource allocation, risk, environmental quality and land use. The crop growth model in EPIC requires daily data on maximum and minimum temperature, precipitation, solar radiation, relative humidity and wind speed. Long-term records of daily temperature and precipitation

Thank you for your answers Dear Mr. Hossein Zare &  Mr Hussain Falih Mahdi

The question is what do you want to know? the crop production in terms of above ground biomass or the crop adaptation?

In every case the weather variables that you have (forecasted) in the future climate change, affect the procedure and models that you can apply.

Just for example: if you have only the temperatures and rainfall (this is very diffuse case in downscaling approach), you can apply a model (hydrological or agro-hydrological) to solve the soil water balance and after to estimate the biomass production from actual transpiration on the base of concept of water productivity (water driven growth engine. E.g. ACQUACROP).

In alternative, with these weather information you can make an estimation of crop adaptation to climate change using the output of simulation model (agro-hydrological model..e.g. SWAP) as an hydrological indicator (e.g. Relative evapotraspiration deficit) that have to be compared with a specific crop hybrid/cultivar threshold for adaptation (this threshold is obtained from the relative yield functions) (see Menenti et al. 2014 "Adaptation of Irrigated and Rainfed Agriculture to Climate Change: The Vulnerability of Production Systems and the Potential of Intraspecific Biodiversity (Case Studies in Italy)". In “Handbook of Climate Change Adaptation”. Springer-Verlag Berlin Heidelberg. DOI 10.1007/978-3-642-40455-9_54; Monaco, E., A. Bonfante, S.M. Alfieri, A. Basile, M. Menenti, and F. De Lorenzi. 2014. Climate change, effective water use for irrigation and adaptability of maize: A case study in southern Italy, Biosystems Engineering, http://dx.doi.org/10.1016/j.biosystemseng.2014.09.001).

Naturally, the model that you apply must be calibrated or tested in your case study area. But is very important is that you calibrate the soil water balance, because it is at the base of your analysis and your results (then you need information about soil water content at different depth in time, the pedological description of soils...etc...Hydraulic properties of soil horizons.....)

The alternative to estimate the crop yield production is the use of model that actively simulate the crop growth (below you can find a list of alternatives), but you need the information about the global solar radiation (and also the crop growth parameters estimated by means of calibration in your soil or pedoclimatic area).

In the future climate forecasting is not easy to estimate it, and difficultly you find this information in weather dataset, because you should know the cloudiness. But there is a way to exceed this problem, there is the possibility to estimate the solar radiation with RADEST software (you can find it online) that must be first calibrated on local dataset with solar radiation. (to estimate the parameters of function used to estimate the radiation).

This is the list of models that can be used to estimate the crop production, i hope that my comment will be useful for you.

Nowadays the crop growth modeling determines the dry matter accumulation (or potential production) are based on four principal approaches:

a. The light use efficiency (LUE): It is a simplified approach applied in the LINTUL family models. LINTUL (Light INTerception and UtiLisation) (Spitters and Schapendonk 1990) model uses the linear relationship between biomass production and the amount of radiation intercepted (captured) by the crop canopy (Monteith, 1977), which has been found for many crop species, grown under well-watered conditions and ample nutrient supply, in the absence of pests, diseases and weeds.

b. The photosynthesis approach: It is a process oriented approach where intercepted photoactive radiation and CO2 assimilation are calculated by considering direct and diffusive radiation components, shaded and sunlit leaves and also different spatial layers within the canopy. This concept is applied in the SUCROS (Simple and Universal CROp growth Simulator) family models (Goudriaan and Van Laar, 1994; Van Laar et al., 1997), WOFOST (Boogaard et al. 1998) and INTERCOM (Kropff and van Laar 1993).

c. The radiation use efficiency (RUE): It is used in the CERES model family (Jones and Kiniry 1986; Ritchie et al. 1987) recently, a biochemical model for the photosynthesis process has been implemented in the CERES-Maize model (Lizaso et al. 2005).

d. Water-driven growth engine. Based on the concept of biomass water productivity (de Wit, 1958) rielaboreted by Steduto in AQUACROP model.

The crop growth models of DSSAT (including CERES models mentioned by A. Bonfante) allows a simple exploratory evaluation of the effects of climate change on crops. The user can modify the values of daily weather data by multiplying or adding/subtracting values and run the priviously calibrated/evaluated model and see the effects. One can also evaluate the effectiveness of some crop management practices to cope with climate change effects, such as, the use of no-till, mulching, crop variety with different rooting system depth and so on.

Dear Camilo, On the base of what you want to do and how is your level of knowledge of system that you want to simulate, you can move between the different alternative models to study the crop responses to climate change. But is important that you are able to understand the limitations of your approach and its empiricism (e.g. changing of soil management...is so general and treated in empirics way in the model..). Most of people that use a model forgot the importance of soil water balance and than the soil description. You can have to soils with the same textural class but very different hydraulic properties (e.g the andosols are very particular soils that have to be treated with attention). All depend on your knowledge of system that you want to simulate. The model are not a game, the user must study how they treat the processes of system that you simulate.

Dear Bonfante, that´s what I mentioned a "calibrated/evaluated" model. Tropical soils like in Brazil have a completely different behaviour regarding water retention and we are allware of it! We also know all the simplifications buint in the models! We´ve working with CERES-Maize and CERES-Sorghum for some years. We have already checked for our conditions soil-water balance and nitrate leaching. Of course there are some many other things we need to improve! Maybe we can joint efforts to deep our research!

Dear Zeynab (and dear Bonfante)

The statement seems to be simple, but it focuses to a very wide horizon in apply agricultural meteorology, and its importance is very high nowadays.

First of all, as Dr. Bonfante said, you must consider the following question: What do you want to know? Maybe, the best statement should be: Where do you want to go?

In my opinion, before to find instruments (model, software, ...) you have to consider: What kind of crops you are analyzing? What are the main climate limiting factors for that? And, what are the critical phenological stages, considering the impact or risks for plants?

After that, you have to obtain the climate requirements for defining the crop-climate parameters, in order to adjust and to run the models. On the bibliography you can find very good parameters, but they must be suitable for the considered genotypes and cropping system. For this reason, a good experimental basis is very important.

For example: If the main limiting factor is air temperature (available degree-days, stresses by extreme temperature, or depression in net assimilation). This can be the case of temperate cereals (wheat, oat, barley, ...) or even irrigated tropical cereals, like rice, maize, and others. In this case, the minimum temperatures can be more important than the daily mean temperature. Several studies have shown the high importance of nocturnal air temperature, which are changing faster than diurnal temperature in the context of climate changes. And, frequently this is not considered in crop modeling. Besides, sometimes data of daily minimum temperatures are not available.

In rainfed summer crop conditions (or dry regions) the water deficit is frequently the most important factor, in terms of climate risks. In this case, the explanation of Dr. Bonfante is very clear and suitable. Actually, the water deficit can explain a large portion of variations in maize yields due to climate variability, regardless of temporal tendency. And both empirical or mechanistic models can be use for quantifying those variations. However, even in this case, the use of suitable parameters is indispensable. Moreover, you have to consider the critical periods of crops, in order to improve the efficiency of predictions.

Analyzing a ten-year experimental series for maize, in a wide range of water conditions, we found that the ETr/ETm ratio explained the variations of grain yields (determination coefficient) in 45% for the entire crop cycle, in 50% from tasseling stage to 30 days after, and in 76% for a short period between 2 days before and 7 days after the tasseling stage (Bergamaschi et al. 2006 - http://dx.doi.org/10.1590/S0100-204X2006000200008). So, a simple model can be use in this case if you want to estimate only the grain yield. And estimations can be done several weeks before the crop maturity.

On the other hand, if you want to simulate other parameters of crop growth, besides of grain yield, you must use mechanistic models. In this case, you will need a large set of soil and crop parameters, and you have to consider the recommendations given above. We have did it with the Glam-Maize model, which was adjusted for subtropical conditions. You can find it (in English) in the following paper: Bergamaschi et al. 2013 - http://dx.doi.org/10.1590/S0100-204X2013000200002.

Very nice intervention! Simple models in many cases work better then those complex models like DSSAT!

Thanks, Camilo.

Your statement was important for remembering a very important aspect of modeling.

My intention was not to compare simple and complex models, because they have different characteristics and applications. From this point of view the comparison doesn't make sense. Besides, if you are using a large and reliable set of experimental data to adjust an empirical model it became, in some sense, also complex. Anyway, it remain as an empirical model and, therefore, it should be reliable only for a narrow and limited range of conditions.

I referred some relationships between the ETr/ETm ratio and maize yields just for emphasizing the importance of precise and reliable crop parameters, regardless of the model complexity. In this case, the intention was to enhance the effect of water deficit during the critical period, which comprises pollination, fecundation, and grain setting. Considering that, maybe the importance of experimental support should be higher for empirical than for mechanistic models.

Am I wrong?