GEOG60941: Environmental Remote Sensing 2024-2025
Assessment 1: Long term monitoring of the Earth’s surface
Deadline: 14:00 Thursday 7th November (week 7). Submission via the Assessment Submission tab on the BB site.
This assessment is worth 20% of the overall mark. The total word limit for this assessment is 500 words. This includes all text but excludes the assessment title, reference list, numbers in tables and text that is included in figures and figure captions. Everything else is included.
Task:
Create a poster with the title “Remote sensing as a tool for long term quantitative monitoring of land cover change: Opportunities and Challenges.”.
Your answer should discuss the question statement with specific reference to both the literature on the topic AND results obtained from your own data analyses. You should use what you learn in the practical sessions to generate data/outputs to graphically demonstrate key points made within your discussion i.e. refer to your own data as examples or case studies (think about the specific points mentioned below!).
Instructions. Please read them very carefully:
● Make sure that your poster addresses at least the following:
o The role of spectral indices in long term quantitative monitoring of land cover change
o How and why the atmosphere may influence spectral indices and ways in which these issues may be minimised
o The implications of using different sensors to monitor quantitative changes in land cover over time
o Wider discussion of how such challenges are being/can be addressed
o Conclusions with reference to the usefulness of remote sensing for long term quantitative monitoring of land cover change.
● Include a URLlink to your fully commented Javascript code somewhere on your poster (see end of this document for further instructions)
● Any figures or tables included in your answers need a proper figure number and caption and should be referenced from the text.
● You must use references from the literature to back-up your key points in each of your answers. The references (both in text and in the reference list) need to be formatted appropriately. As this is a poster assessment you are allowed to use the numeric referencing system to save space.
● A complete list of references used in all of your answers should be included on your poster
● Do not screen grab images or charts from Google Earth Engine (GEE). All images should be exported from GEE and where necessary, tidied up (e.g. add legends, scale bars and N arrows) in a GIS software or spreadsheet package (e.g. redraw charts in Excel) prior to being inserted into your assessment.
● The poster size should be A1 (the orientation is up to you). An A1 page is 594 x 841mm. You can use a range of software to produce your poster; e.g. Powerpoint, Illustrator, Coral Draw, Inkscape. Whichever software you choose, remember to start by defi ning the document page size; e.g. If you are producing the poster using Powerpoint, you should go to the Design tab, then select Slide Size. Choose Custom Slide Size.
Further help and guidance
Background information
Sentinel-2 is a relatively new platform, which will become increasingly important and prominent in ecosystem studies going forward. However, for exploring historical trends in data, we need to turn to other platforms such as MODIS (since 2000) and Landsat (since the 1970s). There are two main challenges associated with the use of remotely sensed data for quantitative studies of temporal change:
1) The signal recorded by the sensor is not necessarily a good measure of the change in surface conditions (Moran et al. 2001). This is because the signal measured by the sensor is a function of surface reflectance, coupled with sensor calibration, solar zenith angle, sensor viewing angle, and atmospheric influences. In order to be of lasting quantitative value, it is therefore recommended that remotely sensed data be calibrated to physical units such as reflectance.
2) The use of data from multiple sensors is often required to ensure data coverage and continuity, but differences in the spectral characteristics of sensors result in spectral index values being different. Such differences may have implications for long term vegetation monitoring (Cund ill et al., 2015).
References cited
Cund ill, S.L., Van der Werff, H. and Van der Meijde, M., 2015. Adjusting spectral indices for spectral response function differences of very high spatial resolution sensors simulated from field spectra. Sensors, 15(3), pp.6221-6240.
Moran, M. S., R. Bryant, K. Thome, W. Ni, Y. Nouvellon, M. P. Gonzalez-Dugo, and J. Qi. 2001. A refi ned empirical line approach for reflectance factor retrieval from Landsat-5 TM and Landsat-7 ETM+. Remote Sensing of Environment 78:71-82.
Generating data to exemplify points within your answers
I highly recommend that you complete your reading and outline a structure to your poster before embarking upon the practical component of the assessment. This is because you need to know what points you want to exemplify with your data - think about what data you want to display and then work out how to generate those data. Some hints as to what data you could generate to exemplify your points are provided below. Of course you can also generate other data - anything that you want to help with your answer!
● Demonstrate the influence of the atmosphere on vegetation by comparing data from the same satellite products with different levels of atmospheric correction. Think about the impact on the spectral signature(s) but also on any vegetation indices that will be derived from those spectral signatures.
● Demonstrate the potential/limitations surrounding the use of multiple sensors for long term monitoring by processing data from two different sensors for the same location and date (as close as possible!). Think about how the spatial patterns and data values differ between sensors and why?
Whilst most of the image processing should be done using Google Earth Engine, you can also use ArcGIS, QGIS, Excel, R and other packages to further analyse and display your results. It’s important to think about how “best” to create and display examples with a focus on demonstrating your points - you don’t have to create examples using GEE for every point that you make if you don’t deem it to be necessary. You should aim for 3-4 generated outputs.
Hints and tips:
● The bands of Sentinel-2 imagery are different to that of Landsat e.g. cloud cover is controlled by the band ‘CLOUDY_PIXEL_PERCENTAGE’. The spectral bands in Sentinel imagery are also of different spatial resolutions and different to that of Landsat data, which is 30 m spatial resolution. To reduce that level of complexity I have put together some example code for you here to do some additional pre-processing of the Sentinel-2 imagery for you
https://code.earthengine.google.com/5380939cc7e68fb9050c4c70fb393299
● Remember that i) the band positions and wavelengths might differ between different sensors, and ii) pixel size may differ between different sensors; and iii) time of year is an important consideration. For Landsat 9, the metadata property for cloud cover is ‘CLOUD_COVER’ but for Sentinel-2 it is “CLOUD_COVERAGE_ASSESSMENT” (you can get this information from clicking on the name of the image collection and looking at the metadata!)
● There are several “land cover maps” available in GEE if you want to use them. You can search for a land cover map that covers your area of interest and fits your purpose best.
● You can use your knowledge of how to generate spectral signatures/profiles from Practical 1 and 3 if you think a spectral plot may be useful to exemplify your point(s)!
● If you want to “comment out” large chunks of code so that it is not run but not deleted, simply highlight the relevant lines of code and select ctrl+ / to reverse this process simply press ctrl + / again.
● To obtain a link to your GEE script simply click Get Link. A url will appear in your browser address field. You should paste this link at the end of your report (note do not copy and paste the web address from your browser – you must use the Get Link button!). However, if you continue to make edits on this code after the link is generated, they will
not be updated in the linked version so only do this when you are fi nished with the code.
● Remember to use the Google EE help documentation if you are unsure of anything