Mapping coral recovery at Lizard Island

Coral reefs are among the most sensitive ecosystems to climate change. In this tutorial, you’ll explore drone data from Lizard Island, captured before and after the 2024 mass coral bleaching event. You’ll use GeoNadir’s AI and vector tools to measure bleaching extent, coral mortality, and early signs of recovery. This activity demonstrates how remote sensing and AI can transform reef monitoring, replacing hours of underwater survey work with rapid, scalable insights from above.

This activity is based on the following open access research article

Raoult, V., Joyce, K.E. Li, J.Y.Q., Roff, G., Carlin., and Williamson, J.E (2025) Coral bleaching and mass mortality at Lizard Island revealed by drone imagery. Coral Reefs

1. Navigate to https://data.geonadir.com/myprojects

Note that your project interface will look different to this one!

2. Click "New project"

3. Click "Untitled" and change the project name to "Coral bleaching example [[Enter]]"

4. Click the icon to add data.

5. Click "Search existing drone data"

6. Click the "Search" field and type "Lizard [[Enter]]"

7. Click "autel_lizard_northPoint_20240604_1345_20_8580"

When you hover over the name of a dataset, you will see the full name. Make sure that you pick the correct dataset. The name indicates that it was captured on 4 Jun 2024 at 1.45pm and at 20m altitude. The capture date is also listed to the top right of the card.

8. Click "North Point, Lizard Island (northern Great Barrier Reef) hi-res [0.8cm GSD] - 2024 mass coral bleaching event".

Again make sure that you select the correct dataset that has 0.8cm GSD in the name.

9. Click "Add to project"

These datasets were captured before and after the peak of the 2024 bleaching event, allowing us to see immediate post-stress changes.

10. Turn off the more recent June 2024 data to reveal the March data below.

11. You may notice that the March data with all the bleached corals looks a little dull.

We are going to enhance the contrast to fix this. Enhancing contrast helps reveal subtle differences in coral colouration that may indicate bleaching severity, and help us to detect it.

Click the dataset icon next to the layer to open the style bar.

12. Click to expand the contrast enhancement options.

13. Move the red, green, and blue slider bars to a similar location as in the below image.

14. When your image is clearer (personal perception), close the style bar.

15. Turn the June data back on.

16. Select the June layer.

17. Click the Compare layers tool from the top menu bar.

18. Use your mouse to swipe between the layers and explore the differences between the datasets.

19. Press " [[esc]]" to close the Compare layers tool.

20. Download this file of a sample area.

This 10×10 m quadrat sample area mirrors a common field survey size, making your analysis comparable to traditional ecological methods.

21. Click to Add data.

22. Click "Import vectors"

23. Drag and drop the sample area file that you just downloaded into the import layers box (or browse for it).

24. Open the style bar for your sample area layer.

25. Change the style to have 0% fill (it will be transparent) and a black stroke with a weight of 3 so that it's easily visible.

26. Close the style bar.

27. Click the 'kebab' menu for the 10x10 quadrat layer to open more options.

28. Click "Lock / unlock" to lock this layer so that you can't accidentally move or change it while you are working in other areas.

29. Turn off the June data.

30. Open the magic wand tool.

31. Click on a bleached coral.

32. Click "Preview"

33. If you are happy with the result, click "Save"

34. Double-click the name of your new layer "Magic wand" and change to "Bleached coral [[Enter]]"

35. Collect some more samples by clicking on other bleached corals and then click "Preview".

36. Click "Save" if you are happy with the result or otherwise reset and try again.

Each polygon represents an individual bleached coral patch — these can be used to calculate total area and percent cover.

37. Continue with the magic wand until you have captured all of the bleached coral. It should look something like this.

Don't worry if you have polygons that overlap each other or ones that go outside of the 10x10m quadrat. This is ok.

Tip: Take note of the number of polygons and area covered. You can see this in the Table of Contents. In the above example, there are 509 polygons, covering an area of 64.38m^2.

38. Because we have overlapping polygons, this means that we are counting the area of bleached coral more than once. So we need to tidy this up.

With the bleached coral layer selected in the ToC, click on the top menu bar as indicated below to open the cascade of vector editing tools.

39. Click the Dissolve tool.

Dissolve merges overlapping polygons to prevent double-counting.

40. Have another look in your ToC for the number of polygons and area covered.

You should have a single dissolved polygon, and notice that the area covered is smaller now as we are not double counting.

41. However, we are still counting areas outside of our quadrat, so we need to fix that.

With the bleached coral layer selected in your ToC, hold CTRL on your keyboard (CMD on Mac), and also select the 10x10 quadrat layer.

42. From the top menu bar, click the arrow next to the Dissolve icon

43. Click the Intersect tool.

Intersect trims your layer to within the sampling area.

44. Turn off the Bleached coral layer in your ToC so that you can evaluate the Intersect output.

Confirm that it has been 'cookie cut' to the inside of the 10x10 quadrat. You should also notice that the area of the Intersect output is smaller than the bleached coral input now that the edges have been chopped off.

However this tool dissolves all of our polygons again.

45. With your intersection layer selected, return to the top menu bar and click the arrow next to the Intersect tool.

46. Click the Split tool.

While you may be able to see multiple polygons in the map view, they only appear as a single listing in the ToC. The Split tool will break these apart for us.

47. Verify that your polygons have now all been split apart again and you can see the area of each individual polygon. In the example below there are 147 polygons.

48. Turn off the Intersect layer.

49. Turn on the June data layer.

50. Now you should be able to see the more recent data capture.

We are going to repeat the process from above to find the live coral. Unfortunately it won't take very long...

Open the magic wand tool.

51. Zoom in to find one of the few live coral colonies and left click to provide a sample.

52. Click "Preview"

53. If you are happy with the result, click "Save"

54. Double-click "Magic wand" and change the name of the layer to "Live coral June"

55. Find another live coral and collect a sample.

56. Click "Preview"

57. Click "Save"

58. Continue with the magic wand until you have digitised all of the live coral.

59. Now we have the amount of coral that was:

There are a few possible scenarios:

Because we are focused on recovered coral, we are interested in the case where areas were mapped as bleached in March AND alive in June.

Select both the Live coral and Intersected bleached coral layers (remember to hold CTRL or CMD to multi-select).

60. Click the drop down arrow next to Dissolve.

61. Click to Intersect these layers.

62. Double-click "Intersection Live coral June - Intersection Bleached coral - 10x10 quadrat" and change the name to "Recovered coral".

This intersection represents corals that survived bleaching - an indicator of resilience within the population.

63. Open the style bar for Recovered coral

64. Change the fill and stroke to be clearly visible on your map.

65. Often ecologists are interested in the amount of bleached coral per unit area, and will calculate this in the field with 1x1m quadrats. We can do this with drone data too.

Select your final intersected bleached coral layer.

66. From the top menu bar, click to Create heatmap.

67. Change the size of the grid cell from 10 to 1m.

68. Click "Preview"

69. Click "Run"

70. Turn off all the layers above the heatmap.

71. In the ToC, expand the heatmap layer.

72. You will be able to see the legend for the colours on the map. Here the bright pixels are 100% bleached in their 1x1m cell, while the dark pixels have less bleaching.

Click the colour ramp to open the style bar.

73. Click the colour ramp inside the style bar to choose other options.

74. Play with different colour ramps to see what best suits you.

75. Close the style bar when finished.

76. To determine the exact value within any 1x1m cell, open the Inspect tool.

77. Click anywhere on your heatmap to find out the value at that location.

78. Evaluate the results in the Inspect pop up.

Bleaching density maps help identify microhabitats of high stress or resilience, guiding where restoration could focus.