342
A.M.I. Meijerink and C.M.M. Mannaerts
a GIS environment, penmts the combination of spectral infonnation with environmental terrain data such as elevation, soil properties, irrigation channel locations (e.g.,
for seepage) and other environmental factors. Integration of a multi spectral vegetation with soil data and a predictive salinization model, enables early stages of land
degradation and associated productivity losses due to salinization to be detected.
Simple cropping patterns in irrigated land can be mapped using a single image
recorded during the growing season. Agricultural practices and differences in crop
phenology increase the spectral variability of the ground cover in irrigated areas. No
straightforward and simple way of identifying crops with satellite data therefore
exists, except in areas where clear phenological differences between crops strongly
affect the spectral characteristics of the fields. This provides an opportunity to map
cropping patterns in irrigation areas with satellite data.
Besides the use of remote sensing for monitoring in existing irrigation command
areas, high-resolution imagery and aerial photography has been also used for irrigation potential assessment (Moran, 1994). We now observe a trend to view the
management of irrigation schemes in a wider perspective, by considering also the
conditions of the catchment i.e., water supply areas.
Remote sensing images may provide infonnation on changes in land cover, sediment source areas and snowmelt, which may affect water and sediment yield. We
refer to Chap. 17 on Irrigation and Drainage for more discussion on irrigation
efficiency improvement with the aid of remote sensing.
15.7 Decision support systems for water management
15.7.1 Introduction
Many problems in water management are ill-structured because they transcend the
traditional boundaries of the hydrologic sciences. Examples are current hydrologic
and water management issues which need to consider the interactive coupling between
computational hydrology and the environment. Solving these interdisciplinary
problems in water management therefore requires the use of knowledge and expertise
from diverse domains or sources. Knowledge can be of various kinds. Quantitative
knowledge is usually expressed in the fonn of numerical models for simulation of
hydrologic phenomena. Many of the problems in water management have, however,
elements that are not sufficiently defined to be amenable to traditional algorithmic or
numerical techniques. In these cases, a problem-solving environment that is more
robust and relies on a coupling between qualitative reasoning and quantitative
numerical computing has to be explored (Meijerink et al. 1993).
15.7.2 Expert and decision support systems
Expert systems. Although more complex approaches exist (Abbott, 1994), qualitative
knowledge can, in the first instance, be expressed in water management using logical
fonnalisms among hydrologic relevant objects. Objects can be physical, real-world
features such as basin areas, drainage lines or river reaches, landscape units or
A.M.I. Meijerink and C.M.M. Mannaerts
a GIS environment, penmts the combination of spectral infonnation with environmental terrain data such as elevation, soil properties, irrigation channel locations (e.g.,
for seepage) and other environmental factors. Integration of a multi spectral vegetation with soil data and a predictive salinization model, enables early stages of land
degradation and associated productivity losses due to salinization to be detected.
Simple cropping patterns in irrigated land can be mapped using a single image
recorded during the growing season. Agricultural practices and differences in crop
phenology increase the spectral variability of the ground cover in irrigated areas. No
straightforward and simple way of identifying crops with satellite data therefore
exists, except in areas where clear phenological differences between crops strongly
affect the spectral characteristics of the fields. This provides an opportunity to map
cropping patterns in irrigation areas with satellite data.
Besides the use of remote sensing for monitoring in existing irrigation command
areas, high-resolution imagery and aerial photography has been also used for irrigation potential assessment (Moran, 1994). We now observe a trend to view the
management of irrigation schemes in a wider perspective, by considering also the
conditions of the catchment i.e., water supply areas.
Remote sensing images may provide infonnation on changes in land cover, sediment source areas and snowmelt, which may affect water and sediment yield. We
refer to Chap. 17 on Irrigation and Drainage for more discussion on irrigation
efficiency improvement with the aid of remote sensing.
15.7 Decision support systems for water management
15.7.1 Introduction
Many problems in water management are ill-structured because they transcend the
traditional boundaries of the hydrologic sciences. Examples are current hydrologic
and water management issues which need to consider the interactive coupling between
computational hydrology and the environment. Solving these interdisciplinary
problems in water management therefore requires the use of knowledge and expertise
from diverse domains or sources. Knowledge can be of various kinds. Quantitative
knowledge is usually expressed in the fonn of numerical models for simulation of
hydrologic phenomena. Many of the problems in water management have, however,
elements that are not sufficiently defined to be amenable to traditional algorithmic or
numerical techniques. In these cases, a problem-solving environment that is more
robust and relies on a coupling between qualitative reasoning and quantitative
numerical computing has to be explored (Meijerink et al. 1993).
15.7.2 Expert and decision support systems
Expert systems. Although more complex approaches exist (Abbott, 1994), qualitative
knowledge can, in the first instance, be expressed in water management using logical
fonnalisms among hydrologic relevant objects. Objects can be physical, real-world
features such as basin areas, drainage lines or river reaches, landscape units or