---
title: "Dryland assessment - questions of data interpretation"
url: "https://openforis.support/questions/969/dryland-assessment-questions-of-data-interpretation"
tags: ["collectearth"]
status: "unanswered"
score: 0
answer_count: 0
asked: "2016-12-06T16:40:12"
last_activity: "2016-12-06T16:40:12"
source: "Open Foris Support - Q&A Forum"
license: "Copyright Open Foris, 2026. Some rights reserved under creative commons license."
---

# Dryland assessment - questions of data interpretation

Asked by flandsberg on 2016-12-06

Hi!

We are starting to pin down the preliminary card content for our Collect Earth baseline for tree-based landscape restoration in Ethiopia. I read the Dryland Assessment and have a few questions (I originally sent an email to drylands.report@fao.org but it bounced back):

1- how the number of trees (counted individually) relates to % tree. What is the added value of using both? My understanding is that the number of trees should be counted over the whole 0.5 ha plot while the % tree is only counting how many of the dots touch a tree.

2- in Ethiopia, trees can be as small as 2 m height. I am wondering if counting both trees and shrubs makes sense. If someone wants to share his/her opinion, I am interested.

3- the LULC map is based on which card:  land class or IPCC class?

4- when our area crosses dryland and non-dryland, which hierarchical rule to use? Ideally, people would use different hierarchical rules but in practice, it does not seem realistic.

5- Is crown cover density (from the Dryland Assessment) the same as percent tree (in the CE cards)? 

6- the crown cover density map is based on percent tree or number of trees?

7- the number of trees by land use class = average number of trees/ha for a specific land use class x area of that land use class?

8- why the threshold for trees not to be counted is 30?

9- Am I right to think that the map of % tree cover relies on a classification undertaken in GEE with Landsat and uses the spectral value of the central plot point as training area?

10- I realize there are some statistical rules in play that makes that if we have enough points in the area, the plots are representative of the whole area but I am not a statistician and it is hard to grasp. I feel that some of the data collected should not be considered as representative of the whole area but I don't know which ones.... If there are some red flags when interpreting the data from a statistical point of view, thanks for highlighting them.

Cheers and thanks for any help!

Florence


_No answers yet._
