Kilen etale im bōnbōn ko rejejjet ñan bwùni im joñi kein ikkan ko kōjerbal waan ekkàke jiddik ko rej ebbōk pija ko im elap aer alikkar.

Jerbal in etali kein ikkan ko ibwiljin menin eddōk ko jet ej aikuj jimwe im jokin wot juon ilo bōnbōn in kein ikkan ko rejejjet emaroñ pen im lap wōņààn etali ijoko im rabañ ilo kōjerbal kein jerbal ko jej iminene kaki jen iiō ko rellōñ remootlok. Jerbal in ej juōn ekkatak ñan etale elañe ebolel im wōr tokjen kōjerbal waan ekkàke jiddik ko rej pija naetaer (drone) remaroñ kwalok iioo in melele ko jeaikuji ñan kōṃṃani bōnbōn im joñak in kein ikkan ko ilo Aelōñ Kein Ad. Ilo dettan 135 hectare ilo jikin ekkatak eo ion Arno, Aelōñ Kein Ad, drone eo ear ebbōk pija ilo joñan uteej kein ruo 85 im 150 mita jen ion bwūdej im pija kein raar jerbal ñan kōmman màb eo eiio ej àinwōt ñe kwar errelaltak im loe jikin ekkatak in (Orthomosaic), màb eo emaroñ ookoktak im kwalok ijoko jeban kar loi ilo kalimjek juōn pija(3D point clouds), im màb eo ej kwalok uteejen ene emōrà jen ion lojet(digital surface and elevation models). Kajojo joñakin juōn kein ikkan ekoba aetokan im debakbakin ion juōn wōjke, rakar ejaak jen joñak ko būki jen pija ko jen drone eo im keidi ibben joñak ko rijerbal ro raar būki jen jikin ekkatak eo kōjerbal kein joñjoñ ko rej walok ne im mita dettan . Ewōr elōñlok jen roñoul taujin nii ko ralikkar jekar bwūni ilowaan jikin ekkatak eo, innem kemar loe ke joñak in nii ko walok jen joñakin aetokin im debakbakin ion juōn nii kōjerbal pija ko jen drone im joñak ko rijerbal raar loi jen kōjerbal kein joñjoñ ko renañin jokkin wōt juōn kin joñan aetokin (r square value eo 0.83) im debakbakin (r square value eo 0.72). Wōjke ak kein ikkan ko reddik ijoko im relukkun marmari rekar jab lukkun walok aer iioo ilo pija ko jen drone, komman bwe joñjoñ in kein ikkan kein ren jepūkpūk. Melele ko jekar loi jen ad kōjerbal waan ekkàke jiddik kein im rej pija rekwalok aer kapeel ñan kōjbarok jààn ko im būktok joñak im bōnbōn in kein ikkan ko rejejjet im bolel ilo Wōden Bwil kein rettolok, ekoba melele ko reaurok renaj buki jen màb kein.   

Fieldwork

1. Collect sUAS imagery from  85m and 150m altitude
2. Ground - Truthing

• Measure tree height (laser rangefinder)
• Measure tree crown diameter (metric tape)

•Image Processing

•Generate orthomosaics, digital surface and elevation models, and 3D point clouds

•Conduct visible coconut tree count on orthomosaic

•Measure tree height + crown dimensions with ArcGIS and Cloud Compare

• We counted 20,498 coconut trees within the study site.
• We found a good correlation between sUAS-derived metrics and ground base measurements of coconut tree heights with a r squared value of 0.83 and a r squared value of 0.72 for the N-S orientation tree crown diameter. (Refer to images below)

In small remote islands, agroforestry managers are faced with limited financial and human resources to maintain and update inventories. This pilot study demonstrates that sUAS imagery can provide comparable data to traditional ground-based methods, minimizing time spent in the field with reduced costs. In addition, sUAS imagery can also generate useful topographic data. But it is important to note that using these techniques also requires access to geospatial software and computer resources.