Quoted from: https://data.nal.usda.gov/dataset/cpm-cotton-production-model 

Over the last 30 years, the U.S. Department of Agriculture's (USDA) Agricultural Research Service (ARS) has conducted a wide range of research on cotton, including work to develop a series of "production models" designed to serve as decision aids to cotton producers. In 1996, ARS decided to develop a new "second generation" Cotton Production Model (CPM) that would retain the best features of the earlier versions in a new, more versatile, and more user friendly framework. The development process was completed to the stage of beta-testing, when the need to redirect limited resources to other priorities caused ARS to decide not to complete the validation process.

ARS believes that CPM, while only partially validated, has the potential to make useful contributions to American cotton producers when completed. For these reasons, ARS decided to make the model available for further development and commercialization.

The Cotton Production Model (CPM) was developed with a modular structure using an object-oriented programming language, C++. The model draws upon the latest scientific knowledge available, and is intended to be used with a wide variety of cotton types across the entire US Cotton Belt. CPM is written in C++ using a new modular structure that allows flexibility and adaptability. This object-oriented structure should allow modules to be incorporated into process-based models of other crop species (see Acock, B. and V. R. Reddy. 1977. Designing an object-oriented structure for crop models. Ecological Modeling 94: 33-44). In addition to being modular and generic, CPM has other advantages over earlier models. Compared to previous cotton models, CPM is more robust, more user-friendly, more easily maintained, and more easily updated with future advances in science. The algorithms that simulate crop growth are derived in part from the best of each of the previous models, and they incorporate new physiological information as well. A new feature of CPM is that it incorporates 2DSOIL, an excellent up-to-date soil and root process model (see Timlin, D. J., Y. Pachepsky, and B. Acock. 1996. A design for a modular, generic soil simulator to interface with plant models. Agronomy Journal 88:162-169 ). 2DSOIL tracks water movement through the soil-plant-atmosphere continuum with hourly time-steps. It also incorporates a new model of plant water relations that responds realistically to water stress. CPM has updated treatments of carbon and nitrogen stresses compared to previous models, and it is designed for easy addition of responses to phosphorus and potassium. Because the growth of each leaf, inter-node and fruit is simulated separately, CPM should be easily linked to pest or disease models.

CPM has the potential to be useful as a decision aid for cotton farmers and crop production consultants. If fully developed, it would be a valuable tool to optimize management inputs such as irrigation, fertilization, plant growth regulators, and defoliant application prior to harvest. In its current version, however, CPM has not yet been fully validated to be useful as a decision aid. The released version of CPM should be considered an advanced model suitable for research purposes. ARS does not endorse its use for any other purpose at this time. Of particular importance to a decision aid model is the user interface. The interface under which CPM has been developed and tested is one that was earlier developed for the soybean model, GLYCIM, and has been documented elsewhere (Acock, B., Pachepsky, Y. A., Mironenko, E. V., Whisler, F. D., and Reddy, V. R. 1999. GUICS: A Generic User Interface for On-Farm Crop Simulations. Agronomy Journal. 91:657-665). CPM is part of the current release of GUICS.