J. Warren Wilson
The measurement process yields numbers that represent selected characteristics of the system being measured. The process needs to be preceded by clear definition of the characteristics, by choice of units, and by the development of techniques for observation and of instruments to facilitate measurement. Instruments introduce both systematic and random errors; other errors arise through disturbance of the system and through sampling procedures.

Crop research requires measurement of states and of rates (changes in state). Rates do not depend directly on one another, but can be predicted from a knowledge of states and of functional relations. Crops can be studied in terms of structure and composition, and of changes in structure due to physical transport and changes in composition due to chemical conversion. Energy relation are also of basic importance.

Spatial and temporal variability are important aspects of measured values, and lead to the study of patterns and their causes.

It is easier to make measurements on climates than on plants, because plants are more complex and more affected by the measurement process. Where large numbers of measurements can be made readily, data acquisition systems may become necessary.

The value of measurements lies in the relations between them, which may allow exploration of causal relations, testing of hypotheses, and prediction. The complexity of crop/environment relations is such that increasing understanding leads to the need for mathematical models incorporating numerous functional relationships.

Warren Wilson, J. (1973). PLANT AND CLIMATE: MEASURING PROCEDURES AND INSTRUMENTS. Acta Hortic. 32, 129-132
DOI: 10.17660/ActaHortic.1973.32.10