New protected area baseline and option for restoration in future land conservation options #396

pvjeetze · 2022-05-16T14:25:24Z

🐦 Purpose of this PR 🐦

So far MAgPIE only included 351.81 Mha of protected areas (in IUCN Cats. I & II). However, legally protected area in MAgPIE grid cells covered an area of 864.31 Mha in 1995 (IUCN catergories I-VI + not assigned) and 1662.02 in 2020. This implementation closes this gap using data on the protected area in the pasture, primary forest, secondary forest and other land classes estimated based on ESA CCI LULC data and geospatial data on protected areas from https://www.protectedplanet.net/en
The new implementation also allows to restore land to natural vegetation classes based on future conservation templates.

🔧 Checklist for PR creator 🔧

Labeling pull request correctly from the label list.

Low risk : Simple bugfixes (missing files, updated documentation, typos) or Start/output scripts
Medium risk : New realization / Changes to existing realization / Other changes which don't modify default.cfg
High risk : New input files (if cfg$input is changed in default.cfg) / Modification to core model (eg. changes in equations, calculations, introduction of new sets etc.) / Other changes in default.cfg

Providing additional information based on PR label

Low risk : No new model run needed.
Medium risk : Default run based on the current version of the fork from which PR is made
High risk
- Default run from the current develop branch
- Default run based on the current version of the fork from which PR is made

⚠️ Additional notes or warnings ⚠️

New preprocessing has to be set up still

🚨 Checklist for RSE reviewer 🚨

PR is labeled correctly.
CHANGELOG is updated correctly
No hard coded numbers and cluster/country/region names.
No unnecessary increase in module interfaces
All required updates in interfaces (if any) have been properly adressed in the module contracts
In-code comments and documentation comments are satisfactory.
model behavior/performance is satisfactory.
Requested changes (if any) were applied correctly

🚨 Checklist for MAgPIE reviewer 🚨

PR is labeled correctly.
CHANGELOG is updated correctly
No hard coded numbers and cluster/country/region names.
Changes to the model are scientifically sound
In-code comments and documentation comments are satisfactory.
model behavior/performance is satisfactory.
Requested changes (if any) were applied correctly

🔴 lines: default in current develop. 🟢 lines: default of this PR

…f_WDPABaseline

flohump

Thanks for including these improvements!
In general looks fine to me.
I have some smaller comments for clarification.
Consider adding a sentence for the new interface pm_land_conservation in the module contract for 22_land_conservation.
We still need to agree on a new value for cfg$gms$s13_max_gdp_shr.
Disclaimer: I had several bilateral talks with @pvjeetze regarding the suggested model changes in this PR.

CHANGELOG.md

flohump · 2022-05-17T08:37:07Z

config/default.cfg

+cfg$gms$policy_countries22  <- all_iso_countries
+
+# * Whether land restoration is carried out for future options
+cfg$gms$s22_restore_land <- 0                 # def = 0


Why not activated by default?
Spell out all options: 0: off; 1: on

Perfect, I also prefer to have this actually activated by default. This also gives support to suggested changes below of simplifying the condition for restoration in the reference period 1995-2020. If cfg$gms$s22_restore_land <- 1 is the default anyway, we may not need the additional condition of making sure that PAs are restored in the ref period based on the historic data

flohump · 2022-05-17T08:38:18Z

config/default.cfg

+
+# Land conservation policy
+# * Baseline:
+# * (WDPA)          All areas under legal protection from 1995 until 2020


It would be very helpful to have some aggregate global numbers here.
Basically, the numbers you have in the desc of the PR.

flohump · 2022-05-17T08:44:03Z

modules/22_land_conservation/area_based_apr22/realization.gms

+*' areas). Natural vegetation (natveg) and grassland ('past') within protected areas cannot
+*' be converted to other land types. On top of the WDPA baseline protection, there are future
+*' options to protect different conservation priority areas such as biodiversity hotspots (BH),
+*' centers of plant diversity (CBD), Intact Forest Landscapes (IFL) and last of the wild (LW).


Add reference: ", taken from Brooks et al 2006" http://science.sciencemag.org/content/313/5783/58

flohump · 2022-05-17T08:46:28Z

modules/22_land_conservation/area_based_apr22/realization.gms

+*'
+*' @limitations Land cover in the WDPA baseline data is estimated based on ESA-CCI
+*' land-use/land-cover maps from 1995 to 2020, while land pools in MAgPIE are intialised
+*' based on LUH2v2 data. This leads to slight mismatches in some areas. Where the area under


Add that we calibrate the LUH2v2 LU patterns to match FAO numbers for forest.

flohump · 2022-05-17T08:48:36Z

modules/22_land_conservation/area_based_apr22/input.gms

+
+table f22_wdpa_baseline(t_all,j,land) Initial protected area as derived from WDPA until 2020 (mio. ha)
+$ondelim
+$include "./modules/22_land_conservation/input/wdpa_baseline.cs3"


This files is included in "rev4.68_h12_fd712c0b_cellularmagpie_c200_MRI-ESM2-0-ssp370_lpjml-8e6c5eb1.tgz"?
No update of input tgz file vector in default.cfg needed?

True, I will still have to set up a new preprocessing.

I will set up a revision 4.71 today afternoon

flohump · 2022-05-17T08:49:57Z

CHANGELOG.md

@@ -45,6 +47,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - **scripts** Extended dissagregation.R script to replace single "past" land class by LHU range and pastr classes when grassland_apr22 realization is used.
 - **52_carbon** added land carbon sink adjustment factors, needed in R post-processing
 - **core** macros for linear and sigmoidal time interpolation
+- **22_land_conservation** added new module and realisation for land protection. The realisation also includes a new WDPA initialisation data set (from 1995 to 2020) for protected areas under legal protection and meeting IUCN and CBD protected area definitions (including IUCN categories Ia, Ib, III, IV, V, VI and 'not assigned' but legally designated).


Mention the new interface pm_land_conservation

and you can perhaps skip the IUCN categories. The changelog should primarily help to identify what changes happened when, the details should be part of the module/realization documentation

flohump · 2022-05-17T08:53:30Z

modules/22_land_conservation/area_based_apr22/presolve.gms

+* Land restoration
+* -------------------
+
+if(s22_restore_land = 1 OR m_year(t) <= sm_fix_SSP2,


That means irrespective of s22_restore_land there is land restoration between 1995 and 2020?

Can you confirm that this is the intended behavior.

If yes, please clarify in the default.cfg that s22_restore_land has only effects after the year given in sm_fix_SSP2 has passed.
In addition, you could consider introducing an internal switch instead of using sm_fix_SSP2. @tscheypidi What is your suggestion?

I think using sm_fix_SSP2 here is correct as the description of this switch reads " Year until which all parameters are fixed to SSP2 values"

I formulated the conditions as such, as I thought cfg$gms$s22_restore_land <- 0 might become the default. I also think, cfg$gms$s22_restore_land <- 1 would be even better as the default. This would mean I can remove m_year(t) <= sm_fix_SSP2, which made sure that land conservation in the reference period followed the WDPA data

tscheypidi

nice work! I put some comments/change requests. In addition, please also compile the goxygen document and have a look at it. I think it might require some rework / more details in some parts.

tscheypidi · 2022-05-17T09:59:09Z

CHANGELOG.md

@@ -45,6 +47,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - **scripts** Extended dissagregation.R script to replace single "past" land class by LHU range and pastr classes when grassland_apr22 realization is used.
 - **52_carbon** added land carbon sink adjustment factors, needed in R post-processing
 - **core** macros for linear and sigmoidal time interpolation
+- **22_land_conservation** added new module and realisation for land protection. The realisation also includes a new WDPA initialisation data set (from 1995 to 2020) for protected areas under legal protection and meeting IUCN and CBD protected area definitions (including IUCN categories Ia, Ib, III, IV, V, VI and 'not assigned' but legally designated).


and you can perhaps skip the IUCN categories. The changelog should primarily help to identify what changes happened when, the details should be part of the module/realization documentation

tscheypidi · 2022-05-17T10:02:29Z

config/default.cfg

+# * (WDPA)          All areas under legal protection from 1995 until 2020
+# *                 (IUCN categories Ia, Ib, II, III, IV, V, VI + 'not assigned')
+# *                 as derived from WDPA (land cover estimated based on ESA CCI)
+# * Future options:


what means "future options"? Do you mean that you put something on top of WDPA for future time steps? If so, I would remove the separation into baseline and future options as this is (at least for me) for confusing than helpful

config/default.cfg

tscheypidi · 2022-05-17T10:13:12Z

config/default.cfg

+# * Start and target year of land conservation option ('c22_protect_scenario').
+# * Defines when full protection based on 'future options' is reached, using a
+# * sigmoidal trajectory. NOTE: This switch is only relevant
+# * if one of the 'future options' (e.g. BH) is chosen.


ok, now I understand why you make this distinction above into future and baseline. Is it correct, that WDPA is always active, no matter what setting one chooses? If so, it might perhaps make sense to have a switch which only containts the scenarios which are put on top of WDPA and have a setting "none" if you want to just run WDPA. This might make it clearer how the different switches interact with each other.

Yes in the current version WDPA is always active, changing the option to "none" would entail quite changes to the way it is set up right now. "WDPA" is simply a set item in the the input data, which is set to zero. Making it quite efficient in the code structure. Changing it to "none" would require a own $ifthen section to allow for this option

I've now done some rework of the input data mainly for easier use during postprocessing, but also to make clearer the difference between the WDPA baseline and future options for conservation in conservation priority areas. protect_area.cs3 is now replaced with consv_prio_areas.cs3 created with the new function calcConservationPriority in mrland. This also allowed me to simplify/rework the code a bit. Please check again.

tscheypidi · 2022-05-17T10:15:40Z

modules/22_land_conservation/area_based_apr22/equations.gms

+*' @equations
+
+*' Natural land conservation
+*' Total natural land cannot be smaller than total natural land conservation target


Please use whole sentences for the goxygen document. And please also run goxygen to check how the documentation of the new module looks like

tscheypidi · 2022-05-17T10:18:21Z

modules/22_land_conservation/area_based_apr22/equations.gms

+                       =g=
+                       sum(ct, p22_min_forest(ct,j2));
+
+ q22_min_other(j2) .. vm_land(j2,"other") =g= sum(ct, p22_min_other(ct,j2));


why is that a separate constraint? Can't it be part of q22_natveg_conservation?

In addition, the documentation comments could be a bit more detailed/explanatory here.

tscheypidi · 2022-05-17T12:03:04Z

modules/22_land_conservation/area_based_apr22/equations.gms

+ q22_natveg_conservation(j2) ..
+            sum(land_natveg, vm_land(j2,land_natveg))
+            =g=
+            sum((ct,land_natveg,consv_type), pm_land_conservation(ct,j2,land_natveg,consv_type));


Currently, it seems that you have a mix of constraints, some of which are set here in the land conservation module and others which are set in the corresponding land modules (based on pm_land_conservation. Please choose only one of the two options: Either do all constraints here (in which case pm_land_conservation is not needed anymore) or do all in the corresponding land modules (in which case you don't need vm_land anymore as input).

I don't have a clear preference, could only imagine that in some cases the treatment of land conservations constraints might be land realization specific in which case option two would be preferable.

I can easily move q22_natveg_conservation(j2) to the natveg module. Adding NPI/NDC to the land conservation module was also something that I discussed with Florian.. @flohump what would you say? Shall I move the NPI/NDC equation back to the natveg module?

Indeed, we discussed this previously. But looking at the code again, I see Jan's point. It's better to have the constraints in one module, i.e. moving q22_natveg_conservation back to the natveg module. With NPI/NDC I'm a bit torn because content-wise it fits to the new 22_land_conservation module. But reading-in the NPI data in 22_land_conservation module and protecting the area in 35_natveg would require another interface. Therefore, we could also move back the NPI/NDC code to 35_natveg. That means the new module 22 would only provide the interface pm_land_conservation without having any constraints.

tscheypidi · 2022-05-17T12:10:19Z

modules/22_land_conservation/area_based_apr22/presolve.gms

+* Land restoration
+* -------------------
+
+if(s22_restore_land = 1 OR m_year(t) <= sm_fix_SSP2,


I think using sm_fix_SSP2 here is correct as the description of this switch reads " Year until which all parameters are fixed to SSP2 values"

modules/22_land_conservation/module.gms

Co-authored-by: Florian Humpenöder <humpenoeder@pik-potsdam.de>

Co-authored-by: Jan Dietrich <dietrich@pik-potsdam.de>

…f_WDPABaselineNew

pvjeetze · 2022-05-23T10:41:22Z

I've updated the code, please see whether, I have addressed all requests. The updated preprocessing will follow this afternoon.

tscheypidi

Changes look good to me. Just some additional questions / change requests.

modules/22_land_conservation/area_based_apr22/preloop.gms

modules/22_land_conservation/area_based_apr22/presolve.gms

modules/22_land_conservation/module.gms

modules/35_natveg/dynamic_feb21/declarations.gms

modules/35_natveg/dynamic_feb21/equations.gms

flohump

I looked again through all changes and did not find any inconsistencies.

tscheypidi

looks good

pvjeetze added 30 commits April 12, 2022 10:48

added new module 22_land_protection

2f4ea44

resolved merge conflicts

022765a

updated module and realisation descriptions in 22 and 35

194523f

replaced fader files with macros in modules 22 and 35

da01299

bugfix

ad90834

bugfix

487b79f

bugfix

9fbd96d

bugfix

5d2ebfe

minor brush-ups

08e3b9b

bugfix

36160ba

bugfix

1c76353

minor brush-ups

672bd3f

added land restoration

5dca3f0

changed restoration equation

9a0a0ea

restoration in 35 presolve

8d26fb8

changed secdforest restoration

30d15eb

only other restoration

b919887

only secdforest restoration

2a9869f

refined switches and updated scenario config

b538424

pasture restoration

6319969

Merge branch 'develop' of https://github.com/magpiemodel/magpie into …

1aa7724

…f_WDPABaseline

update

b4e6d7e

fix for pasture restoration

3314b96

fixed other land restoration

4fe06ad

some brush-ups

3c3a88d

resolved conflict between NDC afforestation and restoration

9693272

full restoration

ac3995c

slightly changed land restoration switch

dc3588d

land restoration also depends on 'aff' afforestation

a17dafa

fix for forest protection

27f1748

pvjeetze added Major Substantial modifications High risk Higher risk labels May 16, 2022

pvjeetze requested review from tscheypidi, flohump, FelicitasBeier and weindl May 16, 2022 14:25

flohump reviewed May 17, 2022

View reviewed changes

tscheypidi requested changes May 17, 2022

View reviewed changes

pvjeetze and others added 7 commits May 18, 2022 10:35

Update CHANGELOG.md

4534e17

Co-authored-by: Florian Humpenöder <humpenoeder@pik-potsdam.de>

Update config/default.cfg

e818a46

Co-authored-by: Jan Dietrich <dietrich@pik-potsdam.de>

updated default.cfg and module description

9064e56

updated limitations description

ead45f0

moved NPI/NDC equations & natveg_conservation equation back to 35_natveg

ec2d5f5

Merge branch 'develop' of https://github.com/magpiemodel/magpie into …

ae39a43

…f_WDPABaselineNew

minor fixes

1ea4382

updated descriptions in modules 22 and 35

919684c

pvjeetze requested a review from tscheypidi May 26, 2022 09:33

pvjeetze added 4 commits May 28, 2022 12:20

improved and simplified input data structure

f1a16a3

updated calib

5e464c6

minor corrections

e9f0d1a

updated scenario config

b73ed3c

tscheypidi requested changes May 30, 2022

View reviewed changes

pvjeetze added 2 commits May 30, 2022 14:16

code simplifications

3f8383c

consv_prio_areas_0.5.mz also transferred to outputfolder

53b6678

flohump approved these changes May 30, 2022

View reviewed changes

small input fix

577a41f

pvjeetze requested a review from tscheypidi May 31, 2022 07:22

tscheypidi approved these changes May 31, 2022

View reviewed changes

pvjeetze merged commit 77d8e22 into magpiemodel:develop May 31, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

New protected area baseline and option for restoration in future land conservation options #396

New protected area baseline and option for restoration in future land conservation options #396

pvjeetze commented May 16, 2022 •

edited by tscheypidi

Loading

flohump left a comment •

edited

Loading

flohump May 17, 2022

pvjeetze May 18, 2022

flohump May 17, 2022

flohump May 17, 2022

flohump May 17, 2022

flohump May 17, 2022

pvjeetze May 18, 2022

pvjeetze May 23, 2022

flohump May 17, 2022

tscheypidi May 17, 2022

flohump May 17, 2022

flohump May 17, 2022

flohump May 17, 2022

tscheypidi May 17, 2022

pvjeetze May 18, 2022

tscheypidi left a comment

tscheypidi May 17, 2022

tscheypidi May 17, 2022

tscheypidi May 17, 2022

pvjeetze May 18, 2022

pvjeetze May 30, 2022

tscheypidi May 17, 2022

tscheypidi May 17, 2022

tscheypidi May 17, 2022

tscheypidi May 17, 2022

pvjeetze May 18, 2022

flohump May 18, 2022

tscheypidi May 17, 2022

pvjeetze commented May 23, 2022

tscheypidi left a comment

flohump left a comment

tscheypidi left a comment

New protected area baseline and option for restoration in future land conservation options #396

New protected area baseline and option for restoration in future land conservation options #396

Conversation

pvjeetze commented May 16, 2022 • edited by tscheypidi Loading

🐦 Purpose of this PR 🐦

🔧 Checklist for PR creator 🔧

⚠️ Additional notes or warnings ⚠️

🚨 Checklist for RSE reviewer 🚨

🚨 Checklist for MAgPIE reviewer 🚨

flohump left a comment • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

tscheypidi left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

pvjeetze commented May 23, 2022

tscheypidi left a comment

Choose a reason for hiding this comment

flohump left a comment

Choose a reason for hiding this comment

tscheypidi left a comment

Choose a reason for hiding this comment

pvjeetze commented May 16, 2022 •

edited by tscheypidi

Loading

flohump left a comment •

edited

Loading