1. Neil Sims, Glenn Newnham, Jacqui England, Carly Green & Alex Held
Good practice guidance for UN SDG Indicator : Proportion of degraded land over total land area
Neil Sims, Glenn Newnham, Jacqui England, Carly Green & Alex Held
CSIRO Land & water
Neil Sims| Team Leader
October 2017

2. UN SDG 15.3.1 Good Practice Guidance
Assist countries in reporting on UN SDG
Proportion of degraded land over total land area
Guidance documents, not a ‘cook book’
‘Living’ documents
Recommendations from current datasets & methods
Sub-indicator and Indicator Guidance
Land cover
Land Productivity
Carbon Stocks, above & below ground (SOC)
Deriving the Indicator
Document provides guidance, and helps to set reporting parameters
Living document based on datasets and methods available at this point in time
GPG for UN SDG | Neil Sims

3. Baseline, Reporting & Review
Baseline from 1 Jan 2000 to 31 Dec 2015
Reporting every 4 years commencing 2018
In line with UNCCD reporting requirements
Review in 2030
Relative to baseline conditions
In context of Land Degradation Neutrality
Baseline*
Reporting
Review
2000
2015
2018
2022
2026
2030
Baseline period can be interpreted differently to assess different metrics
Have to determine baseline degradation extent to assess change in degraded area over reporting period
Success of SDG reviewed in 2030 in context of LDN
GPG for UN SDG | Neil Sims

4. Sub-Indicator 1 – Land cover & land cover change
Ideally use National datasets
Land cover meta language (LCML)
LCML enables land cover classes to be translated from one classification scheme to another – including National schemes
GPG for UN SDG | Neil Sims

5. Sub-Indicator 1 – Land cover & land cover change
Degradation interpreted from transitions from one landcover type to another
Interpreted in National context
This sub-indicator detects large step changes in the landscape – eg land clearing
This table is an example only. Degradation transitions should be interpreted in national context
GPG for UN SDG | Neil Sims

6. Sub-Indicator 2: Land Productivity
Degradation interpreted from changes in Annual Net Primary Productivity (ANPP)
JRC LPD
Plant growth (NDVI or similar)
Annual growth cycles
‘Small integral’ (h)
Area under curve above growing season minimum
Trend, State, Performance
Year
Detects gradual changes
Productivity is plant growth
Plant growth shows annual cycles over the long term. Timesat can calculate statics for each of the annual growth cycle peaks.
Area under growth cycle curve above a growing-season minimum among most highly correlated with ANPP
Day
GPG for UN SDG | Neil Sims

7. Productivity Metric 1. Trend
Slope of trend in ANPP change over time
16 data points from baseline
Mann-Kendall Z for significance
Reporting trend measured from 8 most recent data points
Optional calibration for water availability (WUE)
All data points used in baseline degradation assessment
8 most recent points for each reporting period
GPG for UN SDG | Neil Sims

8. Productivity Metric 2: State
ANPP compared to historical observations for that spatial unit over time
Over time ANPP has a range of values
Baseline degradation:
1. Classify into 10 classes
2. Identify which class the average of 2013,14&15 lies in
3. Areas in lowest 50% of classes potentially degraded
Report periods:
1. Compare average of feporting period ANPP measurements to baseline classes
2. Reduction of two or more classes indicates potential degradation
GPG for UN SDG | Neil Sims

9. Productivity Metric 3. Performance
1 (90th percentile)
0.5
Min
Max
Not degraded
Potentially degraded
ANPP values observed in similar land capability unit
ANPP in a pixel or region compared to pixels with similar ANPP potential
Phenological groups
land cover and climate
Baseline average of
Reporting average of years since last report
GPG for UN SDG | Neil Sims

10. Productivity degradation
Potential degradation is:
1. Significant negative ANPP trend OR
2. Non-significant trend with BOTH
a. Degradation indicated in State assessment
AND
b. Degradation indicated Performance assessment
Support class matrix
GPG for UN SDG | Neil Sims

11. Sub-Indicator 3: Carbon stocks, above and below ground
Degradation indicated by reduction in Soil Organic Carbon (SOC) stocks
Top 30cm of soil profile
To be replaced by Total Carbon Stock analysis when ready
Land cover / unit SOC
IPCC or CCI-LC defaults
SoilGrids 250m or National datasets
Field data, models
Stock change assessment
Land use & management, (tillage, fertiliser),
Emissions (drainage, burning)
Mineral & organic soils
Change in SOC
Assess change between baseline and reporting period SOC
Default/reference SOC stocks from land cover data, or other SOC maps or models
(SoilGrids is a static map so…)
Modify default levels using stock change assessment
Compare change in SOC between reporting and baseline periods
GPG for UN SDG | Neil Sims

12. Change in Soil Organic Carbon
Statistical methods based on confidence intervals of change
Slow rate of change
High uncertainty in many estimates increase potential errors
Threshold based on direction and magnitude of relative change
Regions with >10% average net reduction in SOC
𝑟 𝑆𝑂𝐶 = ( 𝑆𝑂𝐶 𝑡 𝑛 −𝑆𝑂𝐶 𝑡 0 ) 𝑆𝑂𝐶 𝑡 0 ×100
Where:
𝑟 𝑆𝑂𝐶 = relative change in soil organic carbon for spatial feature (%)
𝑆𝑂𝐶 𝑡 0 = baseline soil organic carbon stock for spatial feature (t C ha-1)
𝑆𝑂𝐶 𝑡 𝑛 = soil organic carbon stock at the end of the monitoring period for spatial feature (t C ha-1)
Potential for significant errors in the statistical method
Threshold method arbitrary, but
GPG for UN SDG | Neil Sims

13. Calculating the Indicator
One-Out-All-Out
One-Out-All-Out
Reported as degradation if identified in any one of the sub-indicators
Aggregation
From pixels to regions
Significant when 10% of region is degraded
Identification of false positives and anomalies
Sub indicator
Indicator
Support Class
Land cover
Productivity
SOC
Degraded 1 Y 2 N 3 4 5 6 7 8
GPG for UN SDG | Neil Sims

14. Reporting degradation
Where:
𝐴(𝐷𝑒𝑔𝑟𝑎𝑑𝑒𝑑) 𝑖,𝑛 is the total area degraded in the land cover class i in the year of monitoring n (ha)
𝐴𝑟𝑒𝑐𝑒𝑛𝑡 𝑖,𝑛 is the area defined as degraded in the monitoring year
𝐴𝑝𝑒𝑟𝑠𝑖𝑠𝑡𝑒𝑛𝑡 𝑖,𝑛 is the area previously defined as degraded which remains degraded in the monitoring year
𝑃 n = 𝐴(𝐷𝑒𝑔𝑟𝑎𝑑𝑒𝑑) n 𝑖 𝑚 A(Total)
𝑃 n is the proportion of land that is degraded over total land area
𝐴(𝐷𝑒𝑔𝑟𝑎𝑑𝑒𝑑) n is the total area degraded in the year of monitoring n (ha)
A(Total) is the total area within the national boundary (ha)
GPG for UN SDG | Neil Sims

15. Recommended minimum reporting inclusions
Descriptions and justification of National sub-indicators
Details of data sources and calculations
Land cover legend and transition matrix (maps and justification)
Degradation maps at Indicator and Sub-Indicator level
Justification for “explained anomalies” or “false positives”
Descriptive review of areas identified as degraded
GPG for UN SDG | Neil Sims

16. Reviewers and collaborators
GPG for UN SDG | Neil Sims

17. Thank you CSIRO LAND & WATER
Neil Sims Remote sensing research scientist t e w
Land and water

18. ESA Climate Change Initiative – Land Cover
22 classes, annual between
GPG for UN SDG | Neil Sims

19. Land Productivity Trend (over time)
Metrics correlated with ANPP (NDVI etc)
Performance (relative to other similar land capability units)
All data points used in baseline degradation assessment
8 most recent points for each reporting period
State (relative to past)
GPG for UN SDG | Neil Sims

20. Soil Carbon data IPCC default reference SOC stocks
Some examples of datasets, IPCC default SOC stocks and stock change factors
IPCC default reference SOC stocks
IPCC stock change factors
GPG for UN SDG | Neil Sims