GEIDA training · Uganda · Karamoja

From Uganda's 2024 census to a plan for Moroto district

Four exercises in the browser. You start with all 2,205 subcounties of Uganda, narrow down to Moroto by attributes, then measure access to schools and health centres and screen a road upgrade. The heavy analysis runs on the course server through GeoLibre's own Whitebox tools.

Part	Question	Main tools
0	Where in Uganda should a programme focus?	Select by Expression, Export Selected Features
1	Who lives more than 5 km from a school?	Vector Points To Raster, Euclidean Distance, Reclass, Multiply, Zonal statistics
2	Who lives far from a health centre, and from a proper one?	The same chain, with a filter on facility level
3	What would a Moroto–Tapac road upgrade cross?	Buffer, Slope, Reclass, Zonal statistics (summary and class areas), Select by Location
4	How do I share a result as a map?	Style, Layout Composer plugin

The projects named here are hypothetical; the data and places are real. Results under "What you should see" come from test runs on the course server in October 2026. Small differences are normal.

SETUP

Before you start

About 10 minutes, once. You need Chrome, Edge or Firefox, the access code from your trainer, and Moroto_course_data.zip unzipped on your laptop. Download it from geolibre.terrawatch.net/downloads/Moroto_course_data.zip (3 MB). This guide is also at geolibre.terrawatch.net/guide.

Open GeoLibre

Go to https://web.geolibre.app
Install the Remote Processing plugin

Settings›Manage Plugins›Settings tab

Under Manifest URLs paste the address below and click Add. Close the dialog. The panel should show version 0.5.0 or later; if you installed an older version, remove the URL and add it again.

https://geolibre.terrawatch.net/plugin/plugin.json

Settings › Manage Plugins › Settings tab: paste the plugin address and click Add.
Turn it on and open the panel

Plugins›Installed›Remote Processing, then in the top bar Remote Processing›Server connection and files
Connect to the course server

Panel tab Settings & help

Server
https://geolibre.terrawatch.net

Access code
from your trainer

Measure in metres
leave ticked

Click Save and connect. The status turns green and the panel switches to the Data tab. Note your folder id (for example /data/u-7f3k9q2m1x).

The Data tab after connecting: your folder, your files, and the shared course data.

Check the course rasters

Panel tab ››

You should see three rasters. They are already on the server, so you do not upload them.

Click Add to map to view one. A legend box appears at the bottom left of the map: a colour bar with the value range, or for land cover the WorldCover classes in their official colours. Its × removes an entry.

GeoLibre map of Moroto town land cover with legend — Land cover added from the shared data, in WorldCover colours, with the legend box at the bottom left.

Raster	Contents
`moroto_dem_30m.tif`	Elevation, Copernicus DEM 30 m: 719 m on the plains to 3,079 m on Mt Moroto
`moroto_landcover_10m.tif`	ESA WorldCover 2021, 10 m. Classes: 10 tree cover, 20 shrubland, 30 grassland, 40 cropland, 50 built-up, 60 bare, 80 water, 90 wetland
`moroto_population_2024_100m.tif`	WorldPop 2024 (constrained): people per 100 m cell

Vector files in the zip

Open each with Add Data›Vector Layer and choose the file. They do not need uploading: GeoLibre sends a map layer to the server with each job. The four national boundary files are also in the panel under Shared course data›Vectors: click Add to map.

Districts, subregions, regions and the outline were made by dissolving the census subcounties, so all levels line up exactly. Their pop_total is the sum of the subcounties: it leaves out Bidi Bidi settlement and the five Nakapiripirit subcounties without figures.

File	Contents
`uganda_country.geojson`	Uganda outline
`uganda_regions.geojson`	The 4 regions (Central, Eastern, Northern, Western) with `pop_total`
`uganda_subregions.geojson`	The 17 subregions, Karamoja among them, with `pop_total`
`uganda_districts.geojson`	The 146 districts with `pop_total`, `n_subcounties`, `area_km2`
`uganda_subcounties_census2024.geojson`	All 2,205 subcounties with NPHC 2024 census fields (UBOS)
`moroto_subcounties_census2024.geojson`	The 9 subcounties of Moroto district
`moroto_schools_osm.geojson`	106 schools and kindergartens (OpenStreetMap, Moroto + 10 km)
`moroto_health_facilities_osm.geojson`	31 health facilities (OpenStreetMap), duplicates removed, with a `level` field: HC II, HC III, Hospital
`moroto_roads_osm.geojson`	Roads and paths (OpenStreetMap)
`moroto_tapac_road.geojson`	The 27 km route from Moroto town to Tapac (Part 3)

How to run any Whitebox tool on the server

Every Whitebox step below follows the same pattern:

Open the toolbox and find the tool
Processing›Whitebox Toolbox, type the tool name in the search box, click it.
Untick Run locally (WASM)
Top right of the dialog. If it stays ticked, the tool runs on your laptop and cannot read the server files.
Fill in the inputs
Raster input: set the drop-down to Path and paste a server path. Use Copy path next to a file in the panel's Data tab, or the paths given below. Vector input: pick the layer from the drop-down.
Name the output and run
Type the short name given in the step (for example dist_school) or leave Auto. Click Run and wait for Succeeded. The result appears in your folder in the Data tab.
A Whitebox tool dialog (Reclass): untick Run locally (WASM) at the top, paste the input path, type the values and an output name, then Run.

Paths in your folder

Outputs are saved as /data/<your folder>/<name>.tif. When a later step says "Input dist_school.tif", use Copy path on that file in the Data tab.

PART 0

From Uganda to Moroto

Where in Uganda should an education and health access programme start? Selecting by attributes.

Context

A (hypothetical) Karamoja Education and Health Access Programme must choose one district for its first phase. The evidence is the 2024 census at subcounty level.

Objective

Find the subcounties of Uganda where more than 70% of children aged 6–12 are out of school, then narrow them down to one district for the programme's first phase.

Why: The out-of-school share is the programme's targeting indicator. Starting from all 2,205 subcounties and narrowing step by step shows how a shortlist is built from evidence, and each step can be explained to others.
How: Three attribute selections: the subregion (Karamoja, 113 subcounties), the threshold (above 70%: 67 subcounties), and finally one district.
Why Moroto: Moroto is not the district with the most subcounties above 70%: only 2 of its 9 pass (Tapac 89%, Lotisan 80%), against 18 of 19 in Kotido and 17 of 19 in Kaabong. It is chosen as a pilot that can be run and checked. Its two worst subcounties lie close to Moroto town, a regional hub where a programme office can sit. OpenStreetMap has its schools and health facilities mapped (66 schools, 38 facilities), which Parts 1 and 2 need. It mixes town and remote rural subcounties and includes the slopes of Mt Moroto, which Part 3 needs. With 9 subcounties it stays manageable. Lessons from the pilot then go to Kotido and Kaabong.
Outcome: The 9 Moroto subcounties as the study area for every later part.

Build the context: country, regions, subregions, districts

Add Data›Vector Layer, or the panel's Shared course data›Vectors

Load uganda_country, uganda_regions, uganda_subregions and uganda_districts. Click each one and compare pop_total and n_subcounties: Karamoja has 1.43 million people in 113 subcounties.

Style them as outlines so they work as a frame: open the layer's style (Style tab on the right, or the palette icon in the Layers panel), set the fill opacity to 0 and give each level its own line: thick dark for regions, medium for subregions, thin grey for districts. Drag them to the top of the Layers panel.
Load the national subcounties

Add Data›Vector Layer

Choose uganda_subcounties_census2024.geojson. 2,205 subcounties appear, below the boundary outlines.
Explore the attributes

Click a few subcounties and read pop_total, pct_oos612 (share of children 6–12 out of school), pct_watimp (improved water) and pct_grid (grid electricity). Every field is explained in Census attributes.

Optional: colour the layer by pct_oos612 from its style settings (palette icon in the Layers panel) to see the national pattern.
Select the Karamoja subregion

Edit›Select by Expression...

Select features from
uganda_subcounties_census2024

Expression
["==", ["get", "Subregion"], "Karamoja"]

Modify selection by
Creating a new selection

Click Select features. The dialog reports 113 of 2205 features selected.

Select by Expression on the national subcounties: Karamoja selected.

Expressions use MapLibre's JSON form: ["operator", ["get", "field"], value]. The Expression builder... button helps you write them.
Keep only subcounties with more than 70% out of school

Same dialog

Expression
[">", ["get", "pct_oos612"], 70]

Modify selection by
Selecting within the current selection

Click Select features: 67 selected, and 6 could not be evaluated.
Look at the result

Edit›Zoom to Selection

Zoom to Selection: Karamoja subcounties with more than 70% of children out of school, selected in yellow, with districts and subregions as context.
Choose Moroto and make it a layer

Edit›Select by Expression..., then Edit›Export Selected Features as Layer

Expression
["==", ["get", "District"], "Moroto"]

Modify selection by
Creating a new selection

9 subcounties are selected. The same selection works one level up: ["==", ["get", "District"], "Moroto"] on uganda_districts gives the district outline, and ["==", ["get", "Subregion"], "Karamoja"] on uganda_subregions the subregion. The exported layer is your study area. You can also load moroto_subcounties_census2024.geojson, which is the same.

What you should see

Maps of Uganda and Karamoja — Left: out-of-school share across Uganda, Karamoja outlined in green. Right: the 67 Karamoja subcounties above 70%, the 6 without figures, and Moroto. Computed from the course data; your numbers may differ slightly.

74.6%

of children 6–12 out of school in Karamoja; 20–35% elsewhere

Karamoja subcounties above 70% out of school

subcounties with no census figures (not evaluated)

103,639

people in Moroto's 9 subcounties (census 2024)

The 6 subcounties without figures are Nakapiripirit's five and the Lobule refugee camp. The data's data_note field explains why.

Discuss

Why Moroto rather than Amudat (86% out of school)? Think about data coverage (OpenStreetMap has 66 schools and 38 health facilities in Moroto), the terrain of Mt Moroto for Part 3, and the contrast between town and rural subcounties.
What does a missing value do to a selection? Try ["==", ["get", "pop_total"], null]
Try Filter layer instead of Select features. What is the difference?

PART 1

Who lives far from a school?

How many people in each Moroto subcounty live more than 5 km from a school, and does distance line up with out-of-school rates?

Context

The programme wants to know where new classrooms would bring the most children within walking distance (5 km).

Objective

Measure, for each Moroto subcounty, how many people live more than 5 km from the nearest school, and compare it with the share of children out of school.

Why: If distance explains why children are out of school, new classrooms in the right places will help. If it does not (Lotisan: 80% out of school, yet nobody beyond 5 km), the programme must look at other barriers such as cost, pastoral mobility or child labour.
How: A distance surface from every school, cells beyond 5 km marked, multiplied by the population of each cell, then summed per subcounty.
Outcome: A table and map of people beyond 5 km per subcounty, and a first answer to where distance is the problem (Tapac and Rupa).

Schools (points)→Vector Points To Raster→Euclidean Distance→Reclass > 5 km→Multiply × population→Zonal statistics per subcounty

Load the layers

Add Data›Vector Layer

Load moroto_schools_osm.geojson and moroto_subcounties_census2024.geojson (or use your exported Moroto layer).
Convert the school points to a raster on the population grid

Whitebox tool Vector Points To Raster

Input
the schools layer (drop-down)

Base
Path /data/moroto_population_2024_100m.tif

Zero background
ticked

Output
school_r

Using the population raster as the base puts every later raster on the same 100 m grid, so they can be combined cell by cell.
Distance to the nearest school

Whitebox tool Euclidean Distance

Input
Path to school_r.tif (Copy path in the Data tab)

Output
dist_school

Each cell now holds the straight-line distance in metres to the nearest school.
Mark cells beyond 5 km

Whitebox tool Reclass

Input
Path to dist_school.tif

Reclass values
0;0;5000;1;5000;1000000

Output
far_school

Read the values in groups of three: new value; from; up to. Cells 0–5,000 m become 0, cells beyond 5,000 m become 1.
Count the people beyond 5 km

Whitebox tool Multiply

Input 1
Path to far_school.tif

Input 2
Path /data/moroto_population_2024_100m.tif

Output
pop_far_school

Summarise per subcounty

Panel › (open the section)

Run it three times, with Zones = the Moroto subcounties layer and Also download the table as CSV ticked:

Raster	Field prefix	Read
`dist_school.tif`	`dist_`	`dist_mean`: average distance (m)
`pop_far_school.tif`	`far_`	`far_sum`: people beyond 5 km
Shared `moroto_population_2024_100m.tif`	`pop_`	`pop_sum`: all people (WorldPop)

Compare in a spreadsheet

Open the CSVs. Compute far_sum / pop_sum for each subcounty and set it beside the census field pct_oos612.

What you should see

Map of Moroto subcounties by distance to school — Expected result: share of each subcounty's people more than 5 km from a school, with schools as blue points. Computed from the course data; your numbers may differ slightly.

Subcounty	Mean distance	People > 5 km	Share > 5 km	Out of school 6–12
Tapac	8.5 km	6,659	41%	89%
Rupa	28.5 km	6,240	30%	52%
Nadunget	2.8 km	1,032	6%	58%
Loputuk	3.5 km	601	3%	62%
Katikekile	3.4 km	275	2%	65%
Lotisan	2.1 km	0	0%	80%
Nadunget Town Council	1.6 km	0	0%	60%
North Division	0.6 km	0	0%	19%
South Division	0.8 km	0	0%	20%

About 14,800 people (11% of Moroto) live more than 5 km from a school. Tapac stands out: far from schools and 89% of its children out of school.

Discuss

Rupa's mean distance is 28.5 km, yet only 30% of its people are beyond 5 km. Why? (Large empty areas; people live near the schools.)
Lotisan has 80% out of school but nobody beyond 5 km. Which other barriers matter there?
WorldPop puts about 130,000 people in Moroto; the census counts 103,639. Which would you use, and for what?

PART 2

Who lives far from a health centre?

Which subcounties lack access to a health centre that offers more than basic care (HC III and above)?

Context

The district health office plans outreach and new health centres. Uganda's facility levels run from HC II (basic) through HC III and HC IV to hospitals.

Objective

Find who lives more than 5 km from any health facility, and then from one that offers more than basic care (HC III and above).

Why: Counting facilities hides differences in what they offer. An HC II provides basic outpatient care; HC III and above add services such as maternity care and a laboratory. A subcounty can look well served and still lack proper care.
How: The Part 1 chain run twice: once with all facilities, once with only HC III, HC IV and hospitals, selected by attribute.
Outcome: Two comparable maps. The difference between them shows where upgrading an existing HC II would help most (Katikekile, Loputuk), and confirms Tapac as the worst served.

Load and inspect the health facilities

Add Data›Vector Layer

Choose moroto_health_facilities_osm.geojson and click a few points. The level field was derived from the names, because OpenStreetMap tags every facility in Moroto as "hospital" and several were mapped twice (57 points became 31).
Run the access chain for all facilities

Repeat Part 1 steps 2–6 with the health facilities layer. Use the outputs health_r, dist_health, far_health and pop_far_health.
Select HC III and above

Edit›Select by Expression... on the health facilities layer

Expression
["in", ["get", "level"], ["literal", ["HC III", "HC IV", "Hospital"]]]

Click Select features: 16 of 31 selected. Then Edit›Export Selected Features as Layer.
Run the chain again for HC III and above

Repeat Part 1 steps 2–6 with the exported layer. Use the outputs health3_r, dist_health3, far_health3 and pop_far_health3.

What you should see

Two maps of distance to health facilities — Expected result: requiring an HC III or above raises the gap in Katikekile and Loputuk; Tapac is worst served. Computed from the course data; your numbers may differ slightly.

Subcounty	Share > 5 km, any facility	Share > 5 km, HC III or above
Tapac	61%	75%
Rupa	29%	31%
Lotisan	31%	31%
Katikekile	4%	27%
Loputuk	6%	22%
Nadunget	10%	13%
Town council and divisions	0%	0%

20,700

people more than 5 km from any facility (16%)

30,500

people more than 5 km from an HC III or above (23%)

Requiring a proper health centre reveals gaps in Katikekile and Loputuk that the "any facility" map hides. Tapac is the worst served again.

Discuss

OpenStreetMap is incomplete and needed cleaning. How would you check it against the Ministry of Health facility list?
Straight-line distance ignores Mt Moroto. Cost Distance, with the slope raster from Part 3 as the cost, is the next step.

PART 3

Screening a road upgrade: Moroto town to Tapac

What terrain, land cover and people does a corridor of 1 km on each side of the road cross?

Context

Tapac is the least served subcounty in Parts 1 and 2. A (hypothetical) Moroto–Tapac Rural Access Road would upgrade the existing 27 km track, which runs along the foothills of Mt Moroto.

Objective

Screen the proposed Moroto–Tapac road upgrade: what terrain, land cover, people and facilities lie within 1 km of the route.

Why: Parts 1 and 2 show Tapac as the least served subcounty; a better road would connect it to the schools and health centres of Moroto town. Before any design work, an early screening flags engineering risk (steep slopes), environmental and social sensitivities (grazing land, cropland, settlements) and who would benefit.
How: A 1 km corridor around the road; slope from the elevation model and the share steeper than 15°; the area of each land-cover class; the people and facilities inside the corridor.
Outcome: A short corridor screening summary of the kind that goes into a project concept note.

Road line→Buffer 1 km→Slope, Reclass > 15°→Zonal statistics: class areas and summary→Select by Location

Load the road and the facilities

Add Data›Vector Layer

Load moroto_tapac_road.geojson, plus the schools and health facilities if they are not on the map yet.
Make the corridor

Processing›GeoLibre Toolbox›Vector›Buffer

Input layer
moroto_tapac_road

Distance
1

Units
Kilometers

Engine
Client (Turf.js)

Click Run. A Buffer layer appears. This tool runs in your browser and is instant for one line.

GeoLibre Toolbox › Vector › Buffer: 1 km around the road, run in the browser.
Slope from the DEM

Whitebox tool Slope

Input
Path /data/moroto_dem_30m.tif

Units
degrees

Output
slope
Mark steep terrain (over 15°)

Whitebox tool Reclass

Input
Path to slope.tif

Reclass values
0;0;15;1;15;90

Output
steep15
Land cover in the corridor: area of each class

Panel Data›Zonal statistics

Raster
Shared: moroto_landcover_10m.tif

Statistics
Area of each class (land cover, other class rasters)

Zones
Map layer: Buffer

Field prefix
lc_

Click Run zonal statistics on the server. In one run the corridor gets, for every class, its area in km² and its share in %: lc_grassland_km2, lc_grassland_pct, lc_shrubland_pct, lc_cropland_pct and so on. Click the new Class areas layer on the map to read them, or open the CSV.

Zonal statistics with Statistics = Area of each class and the Buffer as zones.
The 1 km corridor around the road and the Class areas result layer; the zonal statistics form is on the right. Click the corridor to read its km² and % per class.

WorldCover codes are named automatically (10 tree, 20 shrubland, 30 grassland, 40 cropland, 50 built-up, 60 bare, 80 water, 90 wetland). Try the same with Zones = the Moroto subcounties to compare land cover between subcounties.

Slope and people in the corridor

Panel ›, Statistics = Summary, Zones = the Buffer layer

Raster	Field prefix	Read
`slope.tif`	`slope_`	mean, median and max slope
`steep15.tif`	`steep_`	mean = share of the corridor steeper than 15°
Shared `moroto_population_2024_100m.tif`	`pop_`	sum = people in the corridor

Facilities along the road

Edit›Select by Location...

Select the schools that are within the Buffer layer, then do the same for the health facilities.
See where the route climbs

Panel Data, next to slope.tif: Add to map

Click the layer in the Layers panel to open its settings and choose a colour ramp. Steep sections show along the Mt Moroto foothills.

What you should see

Slope map of the corridor and land cover bar chart — Expected result: slope along the Moroto–Tapac corridor (steep sections on the Mt Moroto foothills) and the land cover inside it. Computed from the course data; your numbers may differ slightly.

56.7 km²

corridor around a 27.2 km route

11%

of the corridor steeper than 15° (mean slope 5.8°, max 51°)

35,500

people in the corridor (WorldPop 2024)

16 · 5

schools · health facilities in the corridor (incl. Moroto's 2 hospitals)

Land cover in the corridor	Share
Grassland	55%
Shrubland	31%
Cropland	10%
Built-up	2.5%
Tree cover	1.7%

Discuss

Where are the steep sections? Would a realignment avoid them?
Little cropland is affected, but what does grassland mean in Karamoja (pastoralism, grazing routes)?
Who benefits? Using Parts 1 and 2, how many people in Tapac live beyond 5 km of a school or an HC III that the road would connect to Moroto town?

PART 4

Print a result as a map, with the IsDB logo

Turn the Part 1 result into a map you can put in a report: out-of-school children and distance to school in Moroto. You design the page in the Layout Composer plugin.

Objective

Turn one result into a finished map that can go into a report.

Why: Analysis only informs a decision when others can read it. A map with a clear title, legend, scale bar, data sources and logo is the standard output for reports and appraisal documents, and it is what you hand in for Session 3.
How: Style the result layer, compose the page in Layout Composer, export it.
Outcome: A PNG at 300 dpi, plus a saved layout you can reuse for the other maps.

Install Layout Composer (once)

Settings›Manage Plugins, find Layout Composer in the list, click Install

Close the dialog, then turn it on: Plugins›Installed›Layout Composer. A Layout menu appears in the top bar.
Get the IsDB logo file

It is in the course zip as isdb-logo.png, or download it from geolibre.terrawatch.net/downloads/isdb-logo.png.
Prepare the map

Keep on the map the Part 1 zonal statistics layer for pop_far_school, the schools and uganda_districts; hide the rest with the eye icon. Select the zonal layer, open the Style tab and choose Graduated on pct_oos612, 5 classes, yellow to red. Schools: small dark circles. Districts: thin grey outline, no fill.

Rename the layers from the ⋯ menu in the Layers panel, for example "Children 6–12 out of school (%)" and "Schools (OSM)". The legend uses these names. Then zoom to Moroto district.
Open the composer and add the map

Layout›Open Layout Composer

The page is A4 landscape (change it on the Page tab). Click Add map frame (map icon in the top bar). On the Item tab, under Map View, click Match GeoLibre view or choose your layer in Zoom to layer…. Drag the frame's corners to size it, leaving room on the right.

The frame redraws your vector layers sharply at print resolution. Rasters added from the server (land cover, slope) are not redrawn: for those, set the frame's Content to Capture.
Add title, legend, north arrow and scale bar

Title
Text›Heading, double-click it and type Moroto district: children out of school and distance to school

Legend
Add legend; on the Item tab, set the title and remove entries you do not need

North arrow
Add north arrow, pick a style

Scale bar
Scale bar›Single box

Source note
Text›Caption: Data: UBOS NPHC 2024; OpenStreetMap contributors; WorldPop 2024. GEIDA training, IsDB.

Drag items into place, or type exact positions on the Arrange tab (X, Y, W, H in mm).
Add the IsDB logo

Click Image in the top bar, then Choose image… on the Item tab and pick isdb-logo.png. Keep Fit on Contain (keep ratio) and drag it to a free corner.

Layout Composer: map frame, legend, IsDB logo, north arrow and scale bar on an A4 page. The Layers list on the left shows every item on the page.
Export

Click Export (download icon, top right). Choose PNG at 300 dpi (or PDF), type a file name such as Surname_Map1_schools and click Export. The logo, legend and scale bar are all in the file.

Click the save icon to keep the layout as a .json file; open it later to reuse the same design for Maps 2 and 3.

Land cover legend

For a land-cover map, use + Add manual entry… in the legend's Item tab to add each class with its colour. The class colours are shown in the plugin's map legend box (see Setup).

Discuss

Which map tells the story better: out-of-school share or people beyond 5 km? Make both and compare.
What should every map in a report carry? (Title, legend, scale, north arrow, source, date, author.)

FIELDS

Census attributes

What each field in uganda_subcounties_census2024 and moroto_subcounties_census2024 means.

Figures are from the Uganda Bureau of Statistics, National Population and Housing Census 2024, published in the NPHC 2024 Explorer (subcounty profiles). Counts are persons or households as noted; fields starting with pct_ and pop_dens were derived from them for the course. Field names are 10 characters or fewer so they also work in shapefiles.

Used in the exercises

Subregion, District, pop_total, pop6_12, oos6_12, pct_oos612, data_note. Useful for your own questions: pct_watimp, pct_sanimp, pct_grid, pct_noecce, neet1524p, hh_pdm.

Key 4

Field	Meaning	Unit
`SCCode`	UBOS 6-digit subcounty code (text): district (3) + county (1) + subcounty (2). Join key.
`Subcounty`	Subcounty, town council or city division name (UBOS)
`DCode`	UBOS 3-digit district code (text). Join key.
`District`	District name (UBOS)

Admin 5

Field	Meaning	Unit
`CntyCode`	UBOS 4-digit county code
`County`	County name
`Region`	Region (from first digit of DCode)
`SubregCode`	UBOS subregion code
`Subregion`	Subregion name

Derived 12

Field	Meaning	Unit
`area_km2`	Subcounty area, computed in UTM 36N (EPSG:32636)	km2
`pop_dens`	Population density = pop_total / area_km2	persons/km2
`pop6_17`	School-age population 6-17 = pop6_12 + pop13_17	persons
`pct_noecce`	Share of children aged 3-5 not attending early childhood education = noecce3_5 / pop3_5	%
`pct_noprm6`	Share of 6-year-olds who have not started primary = noprim6 / pop6	%
`pct_oos612`	Share of children aged 6-12 out of school = oos6_12 / pop6_12	%
`pct_oos131`	Share of children aged 13-17 out of school = oos13_17 / pop13_17	%
`pct_age0_4`	Share of population aged 0-4	%
`pct_watimp`	Share of households with improved drinking water = wat_imp / households	%
`pct_sanimp`	Share of households with improved sanitation = san_imp / households	%
`pct_grid`	Share of households lighting with grid electricity	%
`pct_subsis`	Share of households in subsistence economy	%

Population by sex 3

Field	Meaning	Unit
`pop_total`	Total population	persons
`pop_male`	Male population	persons
`pop_female`	Female population	persons

Households 3

Field	Meaning	Unit
`hh_pop`	Household population	persons
`households`	Number of households	households
`hh_size`	Average household size	persons/household

Age groups 12

Field	Meaning	Unit
`age0_4`	Population aged 0-4	persons
`age0_17`	Population aged 0-17	persons
`age6_12`	Population aged 6-12	persons
`age13_18`	Population aged 13-18	persons
`age14_64`	Population aged 14-64	persons
`age15plus`	Population aged 15+	persons
`age15_24`	Population aged 15-24	persons
`age18_30`	Population aged 18-30	persons
`age18plus`	Population aged 18+	persons
`age60plus`	Population aged 60+	persons
`age65plus`	Population aged 65+	persons
`age80plus`	Population aged 80+	persons

Birth registration 3

Field	Meaning	Unit
`br_cert`	Persons registered, with certificate	persons
`br_notif`	Persons registered, with notification	persons
`br_none`	Persons not registered	persons

ICT 6

Field	Meaning	Unit
`inet_male`	Persons aged 10+ who used internet, male	persons
`inet_fem`	Persons aged 10+ who used internet, female	persons
`inet_total`	Persons aged 10+ who used internet, total	persons
`hh_radio`	Households owning a radio	households
`hh_tv`	Households owning a television	households
`hh_comp`	Households owning a computer	households

Information sources 9

Field	Meaning	Unit
`inf_radio`	Households whose main source of information is radio	households
`inf_wom`	Households whose main source of information is word of mouth	households
`inf_phone`	Households whose main source of information is phone calls	households
`inf_tv`	Households whose main source of information is television	households
`inf_meet`	Households whose main source of information is community meetings	households
`inf_inet`	Households whose main source of information is internet/social media	households
`inf_annc`	Households whose main source of information is community announcer	households
`inf_print`	Households whose main source of information is print media	households
`inf_other`	Households whose main source of information is other sources	households

Health 2

Field	Meaning	Unit
`hh_mosnet`	Households owning a mosquito net	households
`p_hlthins`	Persons with health insurance	persons

Employment 8

Field	Meaning	Unit
`unemp15n`	Unemployed, aged 15+	persons
`unemp15pct`	Unemployment rate, aged 15+	%
`unemp1464n`	Unemployed, aged 14-64	persons
`unemp1464p`	Unemployment rate, aged 14-64	%
`neet1524n`	Youth not in employment, education or training (NEET), 15-24	persons
`neet1524p`	NEET rate, 15-24	%
`neet1830n`	NEET, 18-30	persons
`neet1830p`	NEET rate, 18-30	%

Water & sanitation 5

Field	Meaning	Unit
`wat_imp`	Households using an improved main drinking water source	households
`wat_unimp`	Households using an unimproved main drinking water source	households
`san_imp`	Households with improved sanitation	households
`san_unimp`	Households with unimproved sanitation	households
`open_def`	Households practising open defecation	households

Lighting 3

Field	Meaning	Unit
`lt_grid`	Households using grid electricity for lighting	households
`lt_solar`	Households using solar for lighting	households
`lt_solgrid`	Households using combined solar and grid for lighting	households

Economic 2

Field	Meaning	Unit
`hh_subsist`	Households in the subsistence economy	households
`hh_pdm`	Households that benefited from the Parish Development Model (PDM)	households

Education 8

Field	Meaning	Unit
`pop3_5`	Population aged 3-5	persons
`noecce3_5`	Aged 3-5 not attending early childhood care and education	persons
`pop6`	Population aged 6	persons
`noprim6`	Aged 6 not started primary	persons
`pop6_12`	Population aged 6-12	persons
`oos6_12`	Out of school, aged 6-12	persons
`pop13_17`	Population aged 13-17	persons
`oos13_17`	Out of school, aged 13-17	persons

Quality 1

Field	Meaning	Unit
`data_note`	Why a row has no figures or no polygon (blank = complete)

A blank value means the census explorer published no figure for that subcounty; data_note says why. The district, subregion and region layers carry only pop_total, n_subcounties and area_km2.

HELP

If something fails

What happens	What to do
The tool says it could not read the input, or runs very slowly	Untick Run locally (WASM) in the tool dialog.
"Failed: Tool execution failed"	Check the input path (use Copy path), or pick the layer from the list. Check the Reclass values are separated by semicolons.
The result is not in the Data tab	Click ↻ Refresh. If it is still missing, the tool failed: read its message.
Status says Wrong access code or Not connected	Re-type the server address and code in Settings & help, then Save and connect.
My files are gone	You are in a new browser or cleared its data. Under Your folder, open Use an existing folder id and enter your old id.
The corridor looks wrong	Use GeoLibre's Buffer (Part 3, step 2), not Whitebox Buffer Vector, which mishandles long winding lines.
The plugin disappeared	Settings›Manage Plugins›Settings: remove the plugin URL, add it again, and turn it on under Plugins›Installed.

Read the data critically

OpenStreetMap is incomplete and the health facilities were cleaned. WorldPop and the census disagree on Moroto's population. The UBOS boundaries are simplified and overlap slightly along district borders.