Frequently asked questions

This page lists some Frequently Asked Questions (FAQs) for the Trends.Earth tool.

General Questions

Is there a user group I can share experiences with and learn from?

Yes, we recently created a Google group for Trends.Earth users so please join and participate! We aim for this group to be a forum for users to post questions about the tool, methods, and datasets in support of Sustainable Development Goals monitoring. The Trends.Earth team will monitor the group and answer questions as needed, but we’ll get the most out of this community if users support each other by answering questions based on their unique background and experiences. The group will also be used for announcements on tool updates and capacity building efforts.

How can I provide feedback on the tool?

There are three ways to give feedback, emailing the project team, visiting the project site and messaging through the anonymous form or rate the toolbox in the plugins menu of QGIS. The project technical team can address questions through trends.earth@conservation.org. Users can rate the toolbox by opening Plugins in QGIS and selecting Manage and Install Plugins. Select All in the side bar and navigate to trends.earth plugin. Click on trends.earth and rate the toolbox by selecting the number of stars you would like to give the plugin, 5 stars being highly satisfied.

Why is my job stuck in “Pending” status?

When you submit an analysis job, Trends.Earth places it in a queue that processes requests in the order they were received. To ensure fair access to computing resources for all users, each account is limited to running a small number of jobs at the same time (typically three). If you already have several jobs running, any new jobs you submit will remain in “Pending” status until one of your active jobs finishes. Once a slot becomes available, your oldest pending job will automatically start processing. You don’t need to take any action - once your earlier jobs complete, your queued jobs will begin shortly afterward.

Datasets

Note

Refer to the Input Data Used in Trends.Earth section for more information on data sources used in Trends.Earth.

When will you update datasets for the current year?

Trends.Earth uses publicly available data, as such the most up to date datasets will be added to the toolbox as soon as the original data providers make them public. If you notice any update that we missed, please do let us know.

Is there an option to download the original data?

Users can download the original data using the Download option within the toolbox.

Will the toolbox support higher resolution datasets?

The toolbox supports AVHRR (8km) and MODIS (250m) data for primary productivity analysis, and ESA LCC CCI (300m) for land cover change analysis.

In partnership with Apacheta and Auspatious <http://auspatious.com/> and funding from the Global Environmental Facility (GEF), our team is currently developing datasets for land productivity dynamics and land cover at 30 meter scale that will cover the Small Islands Developing States (SIDS)

Can the toolbox support analysis with national-level datasets?

Trends.Earth does have the capacity to import, load and process national-level datasets representing soil organic carbon, land cover and land productivity dynamics, allowing users to take advantage of existing datasets that might be of higher spatial resolution at a national-level than the global datasets that are the defaults in the tool. Please note that working with national-level datasets relies on the local processing capacity of user’s computer, so the time to run analysis using custom datasets will depend on the especifications of user’s machine.

Methods

Note

Refer to the Land degradation and SDG 15.3.1 section for more background on analyses available in Trends.Earth.

Productivity

How does the result provided by state differs from trajectory?

The trajectory analysis uses linear regressions and non-parametric tests to identify long term significant trends in primary productivity. This method however, is not able to capture more recent changes in primary productivity, which could be signals of short term processes of improvement or degradation. By comparing a long term mean to the most recent period, state is able to capture such recent changes.

Land cover

Currently, the land cover aggregation is done following the UNCCD guidelines, but that classification does not take into account country level characteristics. Could it be possible to allow the user to define the aggregation criteria?

Users are able to make these changes using the advanced settings in the land cover GUI so that appropriate aggregations occur depending on the context of your country.

How can we isolate woody plant encroachment within the toolbox?

This can be altered using the land cover change matrix in the toolbox. For every transition, the user can mark the change as stable, improvement or degraded. The transition from grassland/rangeland to shrubland may indicate woody encroachment and this transition can be marked as an indicator of degradation.

Carbon stocks

Why use soil organic carbon (SOC) instead of above and below-ground carbon to measure carbon stocks?

The original proposed indicator is Carbon Stocks, which would include above and below ground biomass. However, given the lack of consistently generated and comparable dataset to assess carbon stocks in woody plants (including shrubs), grasses, croplands, and other land cover types both above and below ground, the Good Practice Guidance published by the UNCCD recommends for the time being to use SOC as a proxy.

Is it possible to measure identify processes of degradation linked to salinization using this tool?

Not directly. If salinization caused a reduction in primary productivity, that decrease would be identified by the productivity indicators, but the users would have to use their local knowledge to assign the causes.

Land degradation outputs

How were the layers combined to define the final land degradation layer?

Performance, state, and trajectory (the three metrics of change in productivity) are combined following a modified version of the good practice guidance developed by the UNCCD (in section SDG Indicator 15.3.1 of this manual a table is presented). Productivity, soil carbon, and land cover chance (the three sub-indicators of SDG 15.3.1) are combined using a “one out, all out” principle. In other words: if there is a decline in any of the three indicators at a particular pixel, then that pixel is mapped as being “degraded”.

Why do I see areas the data says are improving or degrading when I know they are not?

The final output should be interpreted as showing areas potentially degraded. The indicator of land degradation is based on changes in productivity, land cover and soil organic carbon. Several factor could lead to the identification of patterns of degradation which do not seem to correlate to what is happening on the ground, the date of analysis being a very important one. If the climatic conditions at the beginning of the analysis were particularly wet, for example, trends from that moment on could show significant decreases in primary productivity, and degradation. The user can use Trends.Earth to address some of this issues correcting by the effect of climate. The resolution of the data could potentially be another limitation. Trends.Earth by default uses global datasets which will not be the most relevant at all scales and geographies. A functionality to use local data will be added shortly.

All of the sub-indicators are measuring vegetation: how does this contribute to understanding and identifying land degradation?

Vegetation is a key component of most ecosystems, and serve as a good proxy for their overall functioning and health. The three sub-indicators used for SDG 15.3.1 measure different aspects of land cover, which do relate to vegetation. Primary productivity directly measures the change in amount of biomass present in one area, but it does not inform us if that change is positive or not (not all increases in plant biomass should be interpreted as improvement). Land cover fills that gap by interpreting the landscape from a thematic perspective looking at what was there before and what is there now. It does include vegetation, but also bare land, urban and water. Finally, the soil organic carbon indicator uses the land cover map to inform the changes in soil organic carbon over time. This method is not ideal, but given the current state of global soil science and surveying, there is consensus that it this point in time and globally, this is the best approach.

Future plans

When will there be an offline version of the toolbox?

The toolbox is available as both as an offline and online version. The online version allows users to access current datasets more easily, while also allowing users to leverage Google Earth Engine to provide computing in the cloud. The offline version allows users to access data and perform analyses where internet connectivity may be limited, but it does have the disadvantage of requiring users to have enough local computing capacity to run analyses locally.