Two types of classification techniques are present: supervised image classification technique and unsupervised image classification technique. In contrast, in the unsupervised after, the classification user can categorize the classified image. In this particular study supervised classification technique is used to classify the 2000 and 2022 images. Both images are classified with the help of a supervised classification technique. Various steps are followed to classify the image.

A categorization method that specifies the LULC classes was considered when creating the LULC map using satellite imageries. The ideal number of LULC classes depends on the specifications of a particular project for a particular application (Mishra et al., 2020). A total of five classes are considered for the classification, and the table below indicates the classes with descriptions.

Finding out the changes is an important process. This stage was part of the post-classification LULC evaluation. The main input for the rate of change detection procedure was supervised categorized images from 2000 and 2022. After the supervised map's raster map was transformed into a polygon (vector form), the region was determined. We first calculate the Area of each type of land cover for both photographs in order to evaluate the change over time using the below equation,

C = [(At2 — At1)/At1 x 100] (1)

where,

At1 = Area of one type of land cover at t1 time.

At2 = Area of one type of land cover at t2 time (latest)

C = Rate of change occurred in percent.

Accuracy evaluation is a key step in LULC classification. The completed classification may only be trusted and trustworthy after passing accuracy checks since thematic maps made from satellite scenes have a risk of inaccuracy owing to several factors, such as the procedures in classification to retrieve satellite data (Mishra et al., 2020). All the type of accuracy is calculated and reported in the table. The different accuracy is calculated using

Figure 0

There are two categories for the research field. First, in 2000, as seen in Figure 4. Additionally, Figure 5 displays the LULC categorization through 2022. Using images from Landsat 7 and Landsat 8, supervised classification was used to assess the change in LULC, shown in Table 3. The first categorization level includes forests, waterbodies, agricultural, barren land, and settlements. It is based on the LULC classification system.

Over the past 22 years, agricultural land and forest have not changed significantly, but there has been a large decline in bare land and an increase in built-up areas and waterbodies. Urbanization is one of the major threats to the region. Deforestation from various sources is seen in the study area.

Figure 6 depicts an intriguing outcome of changes over the past 22 years. The table lists the amount of territory transferred from one class to another, some of which pose major challenges to development and sustainability. The 2000's classified image shows how the entire region is covered in greenery and much less built-up land and the river is surprisingly seen with less water. The 2022^nd year classified image shows quite the opposite results compared with the year 2000's classified image.

The quantitative evaluation of landscape dynamics is a crucial technique for evaluating the consequences of current land use and a device for figuring out the impact of anthropogenic activities on natural resources. In Figure 6, the LULC modifications that have occurred are visually shown.

The built-up Area and waterbody have dramatically increased in the photograph. A decrease in the percentage of arid terrain may facilitate urbanization. Approximately 68 sq km of once wooded Area have been converted to built-up, indicating an increase in urbanization, it has been noticed. By reducing the forest cover, agriculture has expanded by around 68 sq km, indicating the emergence of additional agricultural fields. A portion of the agricultural Area has also been transformed into developed and arid terrain, a symptom of urbanization.

Accuracy assessment is essential for any decision-making. These three types of accuracy are calculated to see how reliable and well-reliable the data is. According to the accuracy assessment conducted for the year 2000, overall accuracy was 88%. For the year 2022, the overall accuracy is calculated and resulted in 87%. For the users and producers, accuracy formulas are used, and the accuracies are displayed in the below Table 5, Table 6.