4  Corrections

4.0.1 Summary:

This week started with some evolution of remote sensing, exploring how it was discovered and showcasing the first sophisticated picture captured as a tester just two months before Landsat was launched. Some of the topics covered included different sensors, such as the whisk broom and push broom types. The evolution of Landsat from July 1972 to the present day was also discussed. However, the main focus of the day revolved around various corrections in remote sensing, including geometric, atmospheric, orthorectification or topographic, and lastly radiometric correction. Part two of the lectures was on data joining and enhancements

I really find the correction methodology to be highly relevant. For instance, I can think of downloading remotely sensed data covered by clouds—what better way to address this issue than to apply atmospheric correction methodology? Or perhaps working with an image that I need to analyze but find it distorted, so the best solution would be to perform geometric correction. I feel that this is one of the most crucial aspects of remote sensing; essentially, any type of analysis you wish to undertake will require some form of correction to align with your needs.

4.0.2 limitations:

Some of the limitation of correction methodology for remotely sensed data is that EO is widely becoming accepted day by day for disaster sort of intervention carrying out correction methology can result in wasting alot of time, when disater occur there is need for a swift intervention.

4.0.3 Comment:

So now I can refelect based on the comment from the lecture towards the end of the lecture so much sense now, as it stated that most of the correction activities is being handled at the backend now

4.1 Application:

A study of an application of correction which was conducted in the Pearl River Delta, Guangdong Province, China, aimed to determine when atmospheric correction is essential for land cover image classification and change detection using Landsat TM data to identify the best-performing correction methods. The researchers evaluated seven absolute atmospheric correction methods, Dark Object Subtraction (DOS) variations, Dense Dark Vegetation (DDV), Modified Dense Dark Vegetation (MDDV), and Parameterized (PARA)—alongside one relative correction method (Ridge Method), against uncorrected raw Landsat TM images collected between 1988 and 1996.

The paper emphasizes that atmospheric effects, such as scattering and absorption, significantly alter satellite signals, making atmospheric correction a fundamental preprocessing step. However, the necessity of atmospheric correction depends on specific applications, data availability, and analytical approaches.

The study concludes that while all tested atmospheric correction methods improved results, simpler methods like DOS or relative correction (e.g., Ridge Method) are recommended for processing large volumes of Landsat images. These approaches effectively bring multitemporal datasets to a common relative scale, making them practical when accurate surface reflectance is not the primary goal. most Importantly, the methods that enhanced classification and change detection the most did not always provide the most precise surface reflectance estimates.(Seto et al., 2000)

4.2 Reflection

What i have realized is that most of the papers i come across applied this methodology some years back, So i don’t ever think this will be application to my future endeavors but at least knowing about the processes is quite enriching.