Last updated: 2023-01-24

Checks: 7 0

Knit directory: mapme.protectedareas/

This reproducible R Markdown analysis was created with workflowr (version 1.7.0). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.


Great! Since the R Markdown file has been committed to the Git repository, you know the exact version of the code that produced these results.

Great job! The global environment was empty. Objects defined in the global environment can affect the analysis in your R Markdown file in unknown ways. For reproduciblity it’s best to always run the code in an empty environment.

The command set.seed(20210305) was run prior to running the code in the R Markdown file. Setting a seed ensures that any results that rely on randomness, e.g. subsampling or permutations, are reproducible.

Great job! Recording the operating system, R version, and package versions is critical for reproducibility.

Nice! There were no cached chunks for this analysis, so you can be confident that you successfully produced the results during this run.

Great job! Using relative paths to the files within your workflowr project makes it easier to run your code on other machines.

Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility.

The results in this page were generated with repository version ee06b13. See the Past versions tab to see a history of the changes made to the R Markdown and HTML files.

Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated:


working directory clean

Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.


These are the previous versions of the repository in which changes were made to the R Markdown (analysis/threat_assessment_bolivia.Rmd) and HTML (docs/threat_assessment_bolivia.html) files. If you’ve configured a remote Git repository (see ?wflow_git_remote), click on the hyperlinks in the table below to view the files as they were in that past version.

File Version Author Date Message
Rmd ee06b13 Andreas Petutschnig 2023-01-24 workflowr::wflow_publish(“analysis/threat_assessment_bolivia.Rmd”)
html a9e82c8 Andreas Petutschnig 2023-01-23 Build site.
Rmd e613f77 Andreas Petutschnig 2023-01-23 first draft of Bolivia fire threat analysis

Introduction

The following graphics are intented to support a peer-discussion on how publicly available geodata can be used to characterize the threat level of protected areas (short: PAs). The goal of this exercise is to define data products that could support the identification of new intervention sites and thus help KfW and its partners in the project preparation phase. In addition, information about the current portfolio and its relevance to protect the most threatened areas can be derived.

The general idea of this analysis is to match the complete network of existing PAs in countries where KfW is active to the the database of supported PAs by KfW. The same methodology can be used later in the project preparation phase to analyse the existing portfolio together with newly projected intervention sites. The analysis is based on publicly available data from the World Database of Protected Areas (WDPA/IUCN).1 The shown data focuses on suggested areas in Bolivia and contains 11 already funded areas from a total of 152 PAs.

In a first analysis step we focus on habitat destruction in terms of primary forest cover loss. To quantify forest cover loss we utilize data from the Global Forest Watch (Hansen et al, 2013)2. Forest cover loss is defined in their work “as a stand-replacement disturbance, or a change from a forest to non-forest state.”. Loss can either be the result of human activities or natural factors such as droughts, forest fires or hurricanes, amongst others. More information on the interpretation and usefullness of this data as well as suggested further steps to advance the threat assessment are given below in the discussion part. The data also include annual forest fire counts, which are taken from the NASA FIRMS data3.

In the analysis below we will focus on two key outcome indicators:

  • Total forest cover loss: Measures the total sum of loss areas in hectare. This variable is able to identify PAs with the highest primary forest cover loss between 2001 and 2020. The identification of high loss areas can be usefull for targeting areas where we might achieve the largest impact in terms of reducing emissions from deforstation and forest degradation.

  • Relative forest cover loss: Measures the percentage of primary forest cover loss inside a PA compared to its total primary forest area in 2000. The identification of PAs with high relative losses can be relevant from a biodiversity perspective. PAs with high relative losses might be places where large parts of the functional forest habitat is lost. Targeting these areas might not only help to protect the floral biodiversity but also the fauna and humans that inhabit these areas and profit from the local forest ecosystem services.

Map of Threatened Areas

The following map depicts relative and absolute forest cover loss for PAs in KfW supported areas. The size of circles is dependent on the total loss (the bigger the total loss, the larger the circle). The color is dependend on the relative loss (red circles indicate areas with high loss). From a threat perspective big red circles could be especially relevant areas for conservation. This map is interactive meaning that you can zoom into the map to see specific countries in more detail and click on areas to get summary statistics. Furthermore, supported PAs from the current and past portfolio of KfW (blue) are displayed with their actual polygon boundaries as well as all other PAs from the partnering countries (grey).

The overlays in red (PAs Suggested) are the prospective future supported areas.

Lollipop plot

Lollipop Plots can be used to compare supported PAs to non-supported PAs in terms of their treat level. The following figure shows the forest cover loss in suggested PAs. It also shows whether any of these PAs has been or is currently supported by a project from KfW. The database also allows to filter for a specific country or to adapt the observed period (currently 2001-2020) or variable (relative loss instead of absolute). Interactive filtering could be enabled once an interactive solution (Dashboard) is in place.

Crosstables

Crosstables convey the same information about absolute as well as relative loss but on an aggregated portfolio level. They allow us to quickly assess whether our past and current portfolio adresses the most threatened areas. The crosstables group all PAs into different “loss groups” that are displayed in the first column (0-10 hectare, 10-100 hectare, etc.). The second column shows how many PAs fall into each group (in%) from the whole universe of PAs in partnerning countries. The third column shows how many PAs from the KfW portfolio fall into each group, the fourth column shows the suggested PAs. Desirable would be a overrepresentation in high loss areas by KfW. Both, absolute losses (first table) and relative losses (second table) are given below.

Absolute Forst Loss (ha) % of all PAs % of KfW Supported PAs % of Suggested PAs
0-10 20.75 10 4.76
10-100 17.92 0 2.38
100-1,000 19.81 10 11.90
1,000-10,000 22.64 40 40.48
>10,000 18.87 40 40.48
Relative Forst Loss % of all PAs % of KfW Supported PAs % of Suggested PAs
0-2 % 36.79 50 19.05
2-5 % 27.36 50 28.57
5-10 % 16.98 0 26.19
10-20 % 9.43 0 11.90
>20 % 9.43 0 14.29

Lineplots

Lineplots can be used to assess the time-trend of one or several PAs. Their advantage is to display the trend data more detailed which allows users to judge whether a trend is increasing or decreasing and whether it is more likely due to natural phenomena (often huge singular spikes in the trend) or anthropogenic causes (steady, similar trends over all years). The downside of lineplots is that they are not able to compare a large set of PAs because the plots will get fairly messy if too many individual lines are displayed. This is especially troublesome if the trends of different displayed PAs differs considerably. You can isolate trends from on or several PAs of a given project by double-clicking on the project number in the legend. Single clicks on additional projects will allow to compare different projects. Both, absolute losses and relative losses are given in separate figures below.

Interpretation & Discussion

As outline above the GFW methodology defines forest cover loss as a “as a stand-replacement disturbance, or a change from a forest to non-forest state.”. Loss can either be the result of human activities or natural factors such as droughts, forest fires or hurricanes, amongst others. Thus, the data currently does not allow the differentiate permanent loss and conversion from temporary loss due to natural factors. A more in-depth analysis of pre-selected areas is therefore recommended when using the current data.

GFW data can be used to get an assessment of old-growth and primary forests as well as associated carbon emissions. It can indicate which areas had been highly threatened either due to natural causes (hurricanes/fires) or human causes (deforestation/logging/degradation and conversion to agriculture/silviculture). Especially useful in this context is to have a look at the whole trend from 2001 to 2020. Huge spikes in the individual trend data might indicate natural causes such as fires and hurricanes. A quick search with google always helps to confirm this hypothesis or you might look at Hurricane data e.g. on the NOAA website which provides historical data for hurricane tracks. More continuous growth in forest loss is probably due to the conversion of natural forests for agricultural purposes. Again you might look at the map given above and activate the satellite layer to see what might be happening below the detected loss. GFW does not (currently) allow to detect regrowth and regeneration but it is planned to provide that feature soon.

Suggested Improvements

  • Analyze subsequent land-cover in loss-areas. This would be possible e.g. by using Landcover data from the ESA Copernicus mission. It would allow to seperate permanent conversion of forest area e.g. for agricultural purposes from natural losses e.g. due to hurricanes.
  • Aggregate existing loss data on a finer spatial grid or create heatmaps. This would allow to better visualize highly threatened spots on the map. This could be especially relevant to quickly identify high pressure zones in large protected areas.
  • Analyse bufferzones: This would allow to detect areas with high forest loss dynamics in its sourroundings.

sessionInfo()
R version 3.6.3 (2020-02-29)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 18.04.6 LTS

Matrix products: default
BLAS:   /usr/lib/x86_64-linux-gnu/openblas/libblas.so.3
LAPACK: /usr/lib/x86_64-linux-gnu/libopenblasp-r0.2.20.so

locale:
 [1] LC_CTYPE=C.UTF-8       LC_NUMERIC=C           LC_TIME=C.UTF-8       
 [4] LC_COLLATE=C.UTF-8     LC_MONETARY=C.UTF-8    LC_MESSAGES=C.UTF-8   
 [7] LC_PAPER=C.UTF-8       LC_NAME=C              LC_ADDRESS=C          
[10] LC_TELEPHONE=C         LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C   

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] knitr_1.40            plotly_4.10.0         RColorBrewer_1.1-3   
 [4] htmltools_0.5.3       scales_1.2.1          ggsci_2.9            
 [7] leaflet.extras2_1.2.0 leaflet.extras_1.0.0  leaflet_2.1.1        
[10] sf_1.0-9              forcats_0.5.2         stringr_1.5.0        
[13] dplyr_1.0.10          purrr_1.0.1           readr_2.1.2          
[16] tidyr_1.2.1           tibble_3.1.8          ggplot2_3.3.6        
[19] tidyverse_1.3.2      

loaded via a namespace (and not attached):
 [1] fs_1.5.2                bit64_4.0.5             lubridate_1.8.0        
 [4] httr_1.4.4              rprojroot_2.0.3         tools_3.6.3            
 [7] backports_1.4.1         bslib_0.4.0             utf8_1.2.2             
[10] R6_2.5.1                KernSmooth_2.23-20      DBI_1.1.3              
[13] lazyeval_0.2.2          colorspace_2.0-3        withr_2.5.0            
[16] tidyselect_1.2.0        bit_4.0.4               compiler_3.6.3         
[19] git2r_0.30.1            cli_3.6.0               rvest_1.0.3            
[22] xml2_1.3.3              labeling_0.4.2          sass_0.4.2             
[25] classInt_0.4-8          proxy_0.4-27            digest_0.6.29          
[28] rmarkdown_2.16          pkgconfig_2.0.3         highr_0.9              
[31] dbplyr_2.2.1            fastmap_1.1.0           htmlwidgets_1.5.4      
[34] rlang_1.0.6             readxl_1.4.1            rstudioapi_0.14        
[37] farver_2.1.1            jquerylib_0.1.4         generics_0.1.3         
[40] jsonlite_1.8.4          vroom_1.5.7             crosstalk_1.2.0        
[43] googlesheets4_1.0.1     magrittr_2.0.3          s2_1.1.2               
[46] Rcpp_1.0.9              munsell_0.5.0           fansi_1.0.3            
[49] lifecycle_1.0.3         stringi_1.7.12          whisker_0.4            
[52] yaml_2.3.5              grid_3.6.3              parallel_3.6.3         
[55] promises_1.2.0.1        crayon_1.5.1            haven_2.5.1            
[58] hms_1.1.2               pillar_1.8.1            wk_0.7.1               
[61] reprex_2.0.2            glue_1.6.2              evaluate_0.16          
[64] leaflet.providers_1.9.0 data.table_1.14.2       modelr_0.1.9           
[67] vctrs_0.5.1             tzdb_0.3.0              httpuv_1.6.5           
[70] cellranger_1.1.0        gtable_0.3.1            assertthat_0.2.1       
[73] cachem_1.0.6            xfun_0.32               broom_1.0.1            
[76] e1071_1.7-12            later_1.3.0             class_7.3-20           
[79] googledrive_2.0.0       viridisLite_0.4.1       gargle_1.2.0           
[82] workflowr_1.7.0         units_0.8-1             ellipsis_0.3.2         

  1. UNEP-WCMC and IUCN (2022), Protected Planet: The World Database on Protected Areas (WDPA) and World Database on Other Effective Area-based Conservation Measures (WD-OECM) [Online], February 2022, Cambridge, UK: UNEP-WCMC and IUCN. Available at: www.protectedplanet.net.

  2. “Hansen, M. C., P. V. Potapov, R. Moore, M. Hancher, S. A. Turubanova, A. Tyukavina, D. Thau, S. V. Stehman, S. J. Goetz, T. R. Loveland, A. Kommareddy, A. Egorov, L. Chini, C. O. Justice, and J. R. G. Townshend. 2013. “High-Resolution Global Maps of 21st-Century Forest Cover Change.” Science 342 (15 November): 850–53. Data available on-line from: http://earthenginepartners.appspot.com/science-2013-global-forest."

  3. “Justice, C.O., Giglio, L., Korontzi, S., Owens, J., Morisette, J., Roy, D., Descloitres, J., Alleaume, S., Petitcolin, F., & Kaufman, Y.J. (2002). The MODIS fire products. Remote Sensing of Environment, 83:244-262. doi:10.1016/S0034-4257(02)00076-7