1 -- THE RELEVANCE OF CHANGE DETECTION

Looking at an image acquired from a remote sensing instrument, we get a snapshot of vegetation condition that can provide a very useful survey of type distribution, relative productivity, and how natural and manmade patterns interact. In many cases, however, we are particularly interested in studying how productivity and land cover types change over a period of time. Here in the latter quarter of the twentieth century we have been able to make use of the repetitive coverage of remote sensing by implementing a change detection methodology. The time lags can involve days or weeks to capture the effects of events like floods or fire, seasonal comparisons to examine productivity, or most commonly durations of years used for broad operational monitoring.

There are many compelling reasons for concern about vegetation dynamics. Vegetation changes can be related to economic program management, natural resource conservation or research advancement on cause-and-effect relationships between biogeochemical cycles and human activities. Having access to accurate information about changes in vegetation cover helps managers and researchers identify and forecast longer term effects, such as climate change and the creeping costs of extensive land and habitat disturbance.

Localized instances of land cover modification have a global impact because they are repeated frequently worldwide and cumulatively represent a large area. A report by the International Geosphere-Biosphere Program estimates that in the last 300 years (beginning with the waning of the Renaissance, the transition from medieval to modern times), there has been a net loss of approximately 6 million km2 of forest (an area about the size of Australia), a net gain of 12 million km2 in cropland (an area about the size of the USA and Mexico), and a net loss of 1.6 million km2 in wetlands (an area about the size of Alaska). Changes of this magnitude have enormous impacts on global processes, particularly the biogeochemical and hydrological cycles, global primary productivity, and biological diversity.

1.1 Managing Global Resources

As populations have continued to grow and global economies have become more interconnected, the need for monitoring global resources has become vitally important. This is especially true because there is a link between global environment and the harvest, exchange and consumption of resources. Of these resources, vegetation (including agriculture) is one of the most important and changeable in terms of human land use. Thus, use is a critical issue in cover change research -- one worthy of more exploration.

1.1.1 Agriculture -- Changes in the global production of agricultural products can greatly affect market prices for particular products. For example, if 30% of the global wheat crop is damaged by drought and insects, the price for wheat rises in response to reduced supplies relative to demand. Many countries depend on the export of agricultural products to maintain foreign earnings. As the need for foreign earnings increases or the prices of export goods decrease, there is incentive to expand agricultural production which in turn may result in increased land use change and possible environmental degradation. Monitoring crop production and land use can provide information related to these human environmental impacts. Moreover, by monitoring crop status and responding to shortages in essential resources (e.g., water), it is possible to manage reserves to compensate for shortages.

1.1.2 Natural resources -- Management of natural resources is frequently affected by national or local interests. Many times, an action in one country will affect another. For example, air pollution released by an industrial area of one nation may cause acid rain that damages the forests of another nation. The change in vigor of forests can be measured, and from the indications a search for the cause of the change can be more directed and persistent. In another kind of scenario, changes in natural vegetation disrupt energy flow and material cycling which in turn result in adverse impacts on human populations. An example of this would be the removal of forest along river floodways, which tends to increase the sediment load of the stream because of increased erosion, leading to rapid decline in water quality as treatment facilities are overwhelmed.

1.2 Warnings of Impending Problems

Using change detection techniques in concert with predictive models it is possible to forecast the effects of environmental modifications, so that dialogue on policy and response can be initiated in an informed manner before a crisis occurs.

1.2.1 Changing climate -- One topic of much scientific and public interest in recent years is the issue of global climate change and global warming. Scientists know that climate has changed frequently and sometimes dramatically in Earth history, for example, during ice age events. The issue of global climate change is currently centered around the effects of human activity, particularly the burning of fossil fuel and the clearing of land for agriculture and urban development. Generally, global temperature is expected to rise and the frequency of severe weather events to increase as a direct result of human modification of atmospheric composition. The effects of climate change are of particular concern for developing countries and their less diverse economies dependent on natural resources for commodities and subsistence. Small changes in resources, such as rainfall or river levels, could have serious implications for the populations of resource dependent economies.

Using remote sensing, some of the factors that contribute to global climate change, including the effects of forest burning, land conversion and net primary production can be monitored. Remote sensing data can monitor changes as they occur and be used in predictive models assessing the impact of future climate change.

1.2.2 Famine -- Every year famine strikes somewhere in the world, due to both political confusion and natural causes such as drought, floods, or pest infestation. Though difficult to prevent, famine can be lessened if there are some early indicators such as reduced rainfall or crop stress. Although meteorological data are critical for predicting famine, satellite images are also useful to study the vigor of vegetation, and to identify specific areas in setting response priorities. Such early warning systems are in place in Africa to mitigate the effects of famine by starting early food aid to afflicted areas.

1.2.3 Unsustainable resource use -- Another application of change detection is locating areas of unsustainable resource use, or habitat degradation, some of which may lead to problems like famine. Unsustainable land use is becoming more important as the world population grows close to 6 billion. As population increases, so does pressure on the cyclical exploitation of renewable natural resources. In less pressured circumstances, after habitat has been disturbed it has a chance to rebound back to its normal state, such as a rotation of crops and fallow, or logging and reforestation. As land use pressures increase, however, many habitats are not given time to recover. The disruptions of soil compaction, erosion, and invasion of undesirable vegetation ensure long-term if not permanent progression through a declining series of stages. In many cases the land is unlikely to return to anything resembling its natural state. This kind of deleterious outcome was less troubling before man's influence extended so widely across more than 25% of Earth's land surface, and before husbandry practices and energy consumption had obvious negative effects on atmospheric conditions.

Change detection tools allow us to locate ecosystem disturbance and, after monitoring areas in decline, determine what effect the degradation has on surrounding habitats. When use of such tools is combined with knowledge gained from past studies, assumptions can be formulated about particular kinds of processes and trends, and a series of outcomes on the biosphere (including humans) can be predicted. With this kind of information, managers, politicians, and consumers can make more intelligent decisions on land husbandry, conservation, and sustainable use.

prior

toc

next