Master Thesis: From the ground to space
Date:
as of now
Background:
There are myriads of physical and regression models developed over hyperspectral data to estimate crop and soil parameters. However, hyperspectral data with reliable spatial resolution are usually collected using spectroradiometer devices. Therefore, connecting the data sampled using a spectroradiometer to the data captured by multispectral satellite images has been always a challenge for researchers. There are generally two types of problems in this procedure:
- Dealing with the difference in spatial resolutions and
- selecting an effective spectral response function to simulate multispectral data.
Your tasks:
In this research, we would like to focus on the latter, spectral response functions. We would like to find an efficient method by screening through the previous efforts of other scientists to firstly, simulate Landsat 8 and Sentinel-2 images using hyperspectral data, and then apply a physical model (in this case PROSPECT-PRO) on the satellite images for the purpose of leaf nitrogen concentration (N%) estimation. This research consists of one agricultural season of data collection over wheat fields (starting from April 2022), laboratory sample analysis, and computer programming.
Requirements:
- MATLAB, Python, or R
- A huge amount of motivation
Point of contact:
- Ali Mokhtari: ali.mokhtari(at)tum.de
- Prof. Dr. rer. nat. Kang Yu: kang.yu(at)tum.de
- Dürnast 9, 85354 Freising