Background

India is an area where remote sensing has often been used to quantify land cover and land cover changes. The country is large and many areas are not easily accessible for ground field surveys due to the nature of the environment and local infrastructure. Vegetation change is an important variable when studying the changes that took place in this region and the effects of agricultural intensification programs on the local environment. However, reliable data on vegetation cover for this region and time period simply does not exist. Forest change data has been estimated on the district level through formal government forest assessments but recent studies using satellite imagery have shown these to claim up to eight times as much forest area as actually exists (Majumdat,1993). Additionally, past vegetation cover cannot be easily determined through fieldwork. Recent forest estimates on the state level have been estimated using low-resolution satellite imagery, but this data is too coarse for village level purposes and has only been conducted for recent dates (Richards & Flint, 1994). However, an archive of remotely sensed satellite data provides the opportunity to generate micro-scale data beginning in 1972. This period from 1970-1980 is an important one in Indian land use history. One decade earlier government programs began to introduce high-yield seed varieties, which would lead to the complete reformation of agriculture in much of the country, known as the Green Revolution.

What is the Green Revolution?

The Green Revolution is the period of dramatic agricultural intensification that took place in a number of developing countries during the 1960’s-1970’s. It was especially successful in India, where it allowed food productivity rates to outstrip population growth. The success of intensification programs allowed India to fend off the looming threat of mass starvation for its booming population. Agricultural intensification practices were aggressively implemented by government programs targeting specific districts with the goal of national food self-sufficiency.

The major agricultural changes that allowed for such successful increases in productivity were centered around the introduction of high-yield seed varieties. While these varieties allowed for extreme increases in food production, they also required far more inputs and labor. Pesticides, fertilizers, mechanized agricultural equipment, and complex, well-controlled irrigation systems were introduced to many regions in India through direct and indirect government assistance programs (Ford Foundation et al., 1959). Consequently, with this success came many complications. Benefits were not spread equally across geographic regions or social classes and many of these new inputs were responsible for drastic environmental change and degradation in localized regions. Specifically, erosion, soil fertility loss, waterlogging, salinization, deforestation, and water pollution have been some of the well-documented effects of these newly adopted practices (Krishna, 1996).

Why is Gujarat interesting to look at during this period?

The areas most prone to the negative environmental effects of the Green Revolution were semi-arid, temperate environments where, ironically, these agricultural changes were most aggressively implemented in India. The area of eastern Gujarat is just one of these environments (see Figure 1). The region has a pronounced dry season from November through April with periodic droughts every 3-4 years. It is generally a densely populated area, even for India, and includes a wide variety of small and large-scale agricultural and urban land uses.

This area is particularly interesting to research during this period because it is especially prone to the negative environmental side effects of agricultural intensification, such as desertification and soil salinization. Additionally, it is home to some of the largest irrigation projects in the country and includes districts which were early target of government agricultural intensification programs (IADP) (Brown, 1971). Therefore, it is important to understand the impacts of agricultural intensification on overall vegetation cover, and more specifically, the relationship between processes of desertification and deforestation and various land use patterns. In order to do this, vegetation change data is needed on a micro-scale for the region.

What is satellite imagery and why use it?

Remote Sensing is a technique whereby sensors are used to determine the intensity of electromagnetic radiation over specific wavelengths (bands) from surface material. In the field of social science and environmental research, this data is generally acquired by satellite or aircraft and is focused on land cover or other environmental variables on Earth. The values of electromagnetic energy are recorded as digital numbers (DN values) for a specified number of channels (each channel representing a particular wavelength range, or band). These numbers can then be viewed as levels of color intensity on a computer screen and viewed with georeferencing information as images of land cover over specified areas.

Satellite remote sensing is arguably the most accessible and useful means of remote sensing for environmental applications because of its digital format, spatial resolution, and temporal coverage. The digital format of satellite data allows researchers to systematically classify, quantify, and aggregate surface material based on its spectral characteristics. Satellite imagery is much more than interesting photographs; it is a data set of digital values recording the amount of solar electromagnetic radiation reflected from the surface. Data is collected on reflectance in wavelengths beyond the visible spectrum, providing additional information to distinguish between ground cover materials. Collecting vegetation information from the ground can be extremely costly and time consuming when looking at large regions. This is especially true of harsh environments or areas with little access infrastructure. Spatial resolutions of commercially available satellite imagery vary from tens of kilometers to one meter, allowing for detailed detection of surface features in an extremely cost and time effective manner. Additionally, with an archive dating to 1972, one can easily create relatively detailed data sets of ground cover dating back almost 30 years. However, data acquisition is limited by cloud cover, spatial resolution, temporal resolution, satellite lifetime, and ground receiving stations. Additionally, the specific times and dates of data collection are dictated by the orbital mechanics of the individual satellite, and not the specific needs of individual researchers or projects.

Landsat MSS, the sensor used in this study, acquired the earliest commercially available satellite data of earth, dating back to 1972. The sensor obtains data over four bands, covering the visible and near infrared spectrum (see Figure 2 for specific wavelengths). Scenes are 170 km by 180 km with a pixel size, or spatial resolution of 57m by 57m. While this is quite coarse resolution, it has the unique advantage of presenting a uniform data system from 1972 to the present, with a possible data acquisition every 16th day for virtually every area on earth.[1] Therefore, for the purposes of socio-economic or environmental research focusing on dates through the 1970’s, Landsat MSS is the only source of consistent and reliable data available.

Given the broad access to environmental variable data that is possible through satellite data, a range of applications for environmental and social science research exist. One of the most common uses of satellite data is the characterization of green vegetation distribution and abundance. The digital nature of remotely sensed data, as well as the unique spectral characteristics of green vegetation, allows for easy and systematic derivation of quantitative measures of vegetation cover. Vegetation coverage is useful to a variety of researchers, as it can be used as an indicator of environmental degradation or productivity. For example, vegetation data has been used to quantify and monitor tropical deforestation and wetland conversion, forecast agricultural productivity, and document urbanization (Liverman et al., 1998). However, not all vegetation data derived from satellite imagery is equally reliable and useful.

Back to Index

For more information please contact Laurel_Firestone@yahoo.com