This is a very general question. You may want to be more specific in terms of software and scale. Basically you need a time series analysis of the LULC data. Important considerations:
- use comparable data in terms of source, classification but also from the same season (if you are carrying out a study over a longer time period)
- Ideally, also carry out a backcasting study to validate your analysis
- Be aware that environmental disasters (like hurricanes or droughts) or economic shocks (global markets) may disrupt your observed time series, but may also make your future predictions invalid.
In Arc GIS You have to first classify LULC classes of each year images by using NDVI equation ( NDVI=(NIR-Red)/(NIR+Red)) through Arc GIS Raster Calculator and calculate area of each classes year wise ie. for 2000, 2010 and 2017 and then do the trend analysis for future prediction. For detail of the NDVI process you can visit ESRI site of Image Processing. You can do trend analysis through graph in Excel Also .
I study the degradation of some lakes throughout many time periods, and I created a linear regression equation according to lake areas for each year (about 5 periods)
I want to create future maps of the lake in next periods using ArcGIS
we have a function in GIS named Cut Fill, you can use this function. this function create maps which have three classes. they show net gain, unchenge and net losses. any of them sey us our classification layers have how and where changes. of course it should say that this function just compare past and today situation of land uses.
Now I Have a classified LULC from landsat 1988,1998, 2008 and 2018. Based on these results I wants to predict 2030 and 2060 LULC change using ArcGIS 10.5.1
can any one helps with defined procedures and steps. Thanks in advance
Sovan Sankalp , I propose you use the Land Change Modeler (LCM) module of TerrSet. Let's say you have three land use and land cover maps (1990, 2010 and 2020). By inputting two land use and land cover maps (year 1990 and 2010) LCM allows you to determine trends, transitions and changes as a first step. As a second, using Markov chain analysis you can set LCM to predict land use and land cover evolution for the year 2020 to which you have an already established, field verified map. By comparing the output of LCM to the actual 2020 land use and land cover map, you will be able to determine the accuracy of LCM's predictive capacity. One calibrated, you can let LCM predict land use and land cover changes for the year 2040 at a considerable confidence interval. You can fidn several instructive tutorials on youtube for the process.