Remote Sensing and GIS for Prediction/Forecasting

How Remote Sensing and GIS can be used to Predicting Urban Growth…

Remote sensing and GIS integrated with simulation model can be used for forecasting or predicting change in land surface feature. These features can be natural or man-made both can be mapped and predict for future. This predicted information can be used for future plans/monitoring of city.

In Remote sensing one can use temporal and multi spectral satellite images. These satellite image are also provided freely by many organizations. Satellite images can be downloaded from their portal but spatial resolution of satellite images can be moderate to low.

GIS is a versatile tool, which can be used in different applications for monitoring and mapping. In GIS many mathematical and statistical tools can be integrated for analysis current scenario and future prospects. GIS has many functions, which can employ for spatial and neighborhood analysis, as well as spatial database and non spatial database in the form of vector data can be created for spatial and spatial statistical analysis.  

There are several simulation models, which can be used for predicting land use/cover for future.

a. Markov’s model

b. Cellular Automata (CA)

c. CA- Markov’s model

d. Artificial Neural Network based model (Multi Layer Perceptron)

e. Some statistical Method

In these models one can also integrate factor driven based change approach for prediction. 


Here, sharing one example of research, in which Markov Model is used for prediction of urban land use for Kota City. In order to simulate the future urban LULC of Kota, Markov model was used for the land use scenario based on the data of 1999, 2011 and 2016. In the beginning, evaluates land cover changes between two different times and calculates the changes. Based on this transitional probabilities are prepared for each urban class using Markov’s model and used to predict for 2016 and 2030. 


One can also validate the model by predicting LULC. So to assess Model, LULC of 2016 was predicted using land use transition during 1999 to 2011 considering 1999 as base year. The predicted 2016 land use was compared with classified land use of 2016 (prepared using Landsat data of 2016). Thereafter prediction for 2030 was done considering 2011 and 2016 as base input.


Post a comment


Digital Image Processing Tutorial-Image Enhancement Part-I

Introduction of GIS

Height Measurement by Aerial Photographs based on Concept of Parallax