Includes news mentions and reported events (signals).
Export
In short
19 mining permits lie in this area.
Mining footprint (satellite)
74.3 ha of the 546,389 ha (0%) shows mining disturbance on satellite images.[3]11.2 ha of it lies outside known permit boundaries (ESDM). This is not a finding of illegality.
Permits19[1]Kementerian ESDM geoportal (Minerba / Monaresia Wilayah Pertambangan), via legacy Eco Mining Data harvest · signal
Mining footprint (satellite)74 ha[3]Eco Mining Data computation: mining footprint (Cheng et al. 2026 · Maus et al. 2022 · ESMI digitised) × subject; inside and outside current ESDM permit geometries · signal
Permits in protected areas2,029 ha[2]Eco Mining Data computation: current permit geometries (ESDM) × area; overlap with WDPA protected areas · signal
What we don’t know yet (20)
River water quality (pollution index)
Air quality (ISPU, proxy for PM2.5)
Permit change history (ESDM diff) · needs Weekly comparison of the ESDM permit register
The total counts land once where sources overlap (Cheng et al. 2026, then Maus et al. 2022, then ESMI digitised).
Outside known permit boundaries is not a finding of illegality: it can be a gap in the ESDM permit layer, an older or expired permit, or another land use.
Forest loss
Tree-cover loss 2001–2025, where tree cover in 2000 was at least 30 %. Annual satellite data (Hansen / University of Maryland).
Loss means tree cover was removed: clearing, but also plantation harvest or fire. It is not a finding of illegal deforestation, and the data is yearly, not monthly.
Permits
Visible with a free account · Permit list and numbersSign in
Kementerian ESDM geoportal (Minerba / Monaresia Wilayah Pertambangan), via legacy Eco Mining Data harvest · geoportal.esdm.go.id · retrieved 18 Oct 2026 · signal · unknown
Eco Mining Data computation: current permit geometries (ESDM) × area; overlap with WDPA protected areas · computed by Eco Mining Data · as of 18 Oct 2026 · signal · derived from the ESDM permit layer and WDPA (WDPA terms apply)Union of current permit geometries clipped to the subject; intersection with protected-area polygons; geodesic area
Eco Mining Data computation: mining footprint (Cheng et al. 2026 · Maus et al. 2022 · ESMI digitised) × subject; inside and outside current ESDM permit geometries · doi.org · computed by Eco Mining Data · as of 31 Dec 2019 · signal · derived from Cheng et al. 2026 and Maus et al. 2022 (CC BY-SA 4.0), ESMI digitised land and the ESDM permit layerFootprint polygons clipped to the subject (geodesic area), counted once where sources overlap (Cheng > Maus > ESMI); the part inside the union of current permit geometries vs outside known permit boundaries (ESDM). Outside is not a finding of illegality: it can be a data gap or an older permit.
Indonesia administrative regions (seed) · github.com · as of 1 Jan 2025 · confirmed · geometry: CC BY 3.0 IGO; codes: BPS public data (open; attribution: Badan Pusat Statistik)
Hansen Global Forest Change GFC-2025-v1.13 · glad.earthengine.app · as of 31 Dec 2025 · confirmed · CC BY 4.0Hansen, M. C. et al. 2013. High-Resolution Global Maps of 21st-Century Forest Cover Change. Science 342: 850-53. Data: Global Forest Change 2000-2025 (v1.13), University of Maryland.
Needs: Metadata Indikator TPB (Bappenas) + RAD TPB per province/regency
Not connected yet
Trend per indicator (on track / moderate / stagnating / decreasing)
Needs: Yearly series of the axis A indicators
Not connected yet
Mercury / ASGM sites (unconfirmed signals)
Visible with a free account
ASGM signals outside any permit
Visible with a free account
Reclamation gap (disturbed minus reclaimed)
Visible with a free account
Estimated mining effect per target (DiD / synthetic control)
Needs: Control areas (matching) + yearly series
Not connected yet
Permit change history (ESDM diff)
Needs: Weekly comparison of the ESDM permit register
Source is empty
Pollution part of the affected area
Needs: Pollution part of the affected area (water, sediment and air measurements); only the physical footprint is measured today
Not connected yet
Recovery after mining
What the land inside the mining footprint looks like from space, year by year (1990–2025)
Actively mined
Mining ground was still seen in the last few years.
Footprint polygons
16
Ever disturbed
24 ha
Vegetation back
10 ha
Abandoned ≥ 5 years, still bare or water
0 ha
Outside known permit boundaries (ESDM)
10 ha
Figures are sums over all polygons; the status of a group is the status with the largest area. This can be a gap in the ESDM permit layer or an older or expired permit: the layer lists current permits only.
Area by status
Actively mined14 ha 23 %
Managed restoration10 ha 18 %
Not visible at 30 m35 ha 59 %
Key years
First mining
2018
Peak mining
2022
Last mining
2025
Mining stopped
2023 · 3 years ago
Mining ground
Bare ground
Water
Oil palm
Farmland
Tree crops
Shrubs
Forest
Peak mining 5.3 ha in 2022. In 2025: tree crops 41 ha, farmland 17 ha, mining ground 1.3 ha.
Show the numbers
Year
Mining ground
Bare ground
Water
Oil palm
Farmland
Tree crops
Shrubs
Forest
1990
0
0
0
0
34
0
3
22
1991
0
0
0
0
36
0
1
22
1992
0
0
0
0
37
0
1
21
1993
0
0
0
0
54
0
0
5
1994
0
0
0
0
57
0
0
2
1995
0
0
0
0
58
0
0
1
1996
0
0
0
0
55
3
0
1
1997
0
0
0
0
56
3
0
0
1998
0
0
0
0
54
3
0
1
1999
0
0
0
0
54
3
0
1
2000
0
0
0
0
54
3
0
1
2001
0
0
0
0
58
0
Whole footprint polygons, including parts outside this subject.
Status per polygon
Free account · Status per polygon, with confidence, the rule that decided it and its life history, needs a free account.
Satellite land cover, not reported reclamation: ESDM reclamation reports are not used. Sources: MapBiomas Indonesia Collection 5 (30 m, CC BY-SA 4.0), JRC Tropical Moist Forest; footprint Maus et al. 2022 and Cheng et al. 2026 (CC BY-SA 4.0); ESDM MODI permit layer. Method rec-1.0 · computed Oct 19, 2026
Media timeline
What was reported, month by month. Headlines as published; events stay signals until a person confirms them.
Model: comparable agriculture land Rp 9,108/m² minus expected reclamation ≈ Rp 15,809/m² (coal). Reclamation costs more than the land is worth: net −Rp 7,897/m², so its value is 0 (76% of model runs negative).
NJOP and compensation values are converted to market-equivalent and older years CPI-indexed. A fallback value is borrowed from a wider area and has a wider band and a lower grade. Not an appraisal. · landvalue.v1 · API (observations: Tier 1)
Value after restoration
What restoring abandoned mining land that is still bare or water could be worth, in rupiah. Figures for Kabupaten Batanghari.
Model estimate, not an appraisal.
MapBiomas mining-class measure
Every 30 m pixel that MapBiomas ever labelled as a mining pit and that is now abandoned and still bare or water.
Net gain
Rp 233 mP10–P90 Rp 108–364 mgrade D
Net present value over 25 years at 10 % a year.
Land counted
3.4 ha · abandoned, still bare or water
Payback
2 yrs
Gain per rupiah of cost
1.05×
Footprint measure
Land inside the mapped mining footprint polygons that is abandoned and still bare or water.
No abandoned bare or water land in this measure here.
The two measures overlap. They are shown side by side and never added together. tn = trillion, bn = billion, m = million (rupiah). The letter after a figure is its grade (A best, D a rough model estimate); the range is P10–P90.
MapBiomas mining-class measure · Every 30 m pixel that MapBiomas ever labelled as a mining pit and that is now abandoned and still bare or water. 1.4 ha still bare and 2 ha water (pit lakes).
From value now to net gain
Value now → value after restoration, plus what restoration adds for everyone, minus the costs, gives the net gain.
Value now (degraded land)model value−Rp 270 mP10–P90 −Rp 595 m – Rp 372 mgrade D
Value after restorationmodel valueRp 260 mP10–P90 Rp 147–697 mgrade D
The value now is negative: the land is worth less than nothing until it is reclaimed. That is shown but not counted as gain; the restoration costs are counted instead.
+Gain in land valuemodel valueRp 215 mP10–P90 Rp 121–323 mgrade D
+Ecosystem servicesmodel valueRp 165 mP10–P90 Rp 111–235 mgrade D
+CarbonRp 8.87 mP10–P90 Rp 3.82–16.4 mgrade D
+Avoided downstream impactsmodel valueRp 55.5 mP10–P90 Rp 29.8–94.5 mgrade D
−Restoration costsRp 222 mP10–P90 Rp 189–262 mgrade D
Remediation
Shown, not added
Plan income (capitalised)model valueRp 150 mP10–P90 Rp 102–223 mgrade D
Reclamation duty dischargedRp 604 mP10–P90 Rp 389–876 mgrade D
By permit zone
Permit zone
Land
Net gain
Inside an expired or revoked permit: reclamation duty applies
–
–
Inside an active permit
0.3 ha
Rp 2.37 mP10–P90 −Rp 4.27 m – Rp 11.7 mgrade D
Inside a permit with another statusPending, suspended or unknown status.
–
–
Outside known permit boundaries (ESDM)This can be a gap in the ESDM permit layer or an older permit: the layer lists current permits only.
3.1 ha
Rp 230 mP10–P90 Rp 111–353 mgrade D
Who benefits
How the benefits are shared, before costs (median path). Benefit-sharing defaults from the Restoration Studio: 80 % cooperative, 20 % village fund.
LandownerThe gain in land value.Rp 182 mgrade D44 %
CooperativePlan income above the land value and the community share of carbon (80 % stays in the cooperative).Rp 5.11 mgrade D1 %
Village fund20 % of the community share goes to the village fund.Rp 1.28 mgrade D0 %
Government and the publicEcosystem services and avoided downstream impacts.Rp 220 mgrade D53 %
Project developerThe rest of the carbon value.Rp 4.26 mgrade D1 %
Source and grade of every part
Part
Per hectare
Source
Value now (degraded land)model valueToday's use value of the degraded land: comparable land value minus the cost of reclamation. It can be negative. A negative value is shown but not counted as gain, because the restoration costs are counted below.
−Rp 79 m/haP10–P90 −Rp 174 m/ha – Rp 109 m/hagrade D
Eco Mining Data land value: degraded mining land = comparable value − reclamation cost (model value, not floored at zero)
Value after restorationmodel valueThe larger of the comparable land value and the capitalised plan income, never both.
Rp 76.1 m/haP10–P90 Rp 43.1–204 m/hagrade D
The larger of (b) the comparable land value (agriculture/plantation; pit lakes × water factor) and (a) the capitalised plan income
Plan income (capitalised)model valueIncome of a default restoration plan (forest, agroforestry, cassava fields, fish cages in pit lakes) over 25 years, discounted. Used in the value after restoration only when it is the larger of the two.
Rp 44 m/haP10–P90 Rp 29.9–65.4 m/hagrade D
Restoration Studio option economics (Sulawesi Utara price table applied nationally), default plan mix per commodity; present value of operating income over the plan period
=Net gainmodel valueRp 233 mP10–P90 Rp 108–364 mgrade D
Value after restoration minus today's use value (a negative value now counts as zero).
Rp 62.9 m/haP10–P90 Rp 35.6–94.8 m/hagrade D
(value after − today's use value, a negative value counted as 0), realised in year 2 and discounted
Ecosystem servicesmodel valueNon-market value of the restored land, such as water regulation and habitat (benefit transfer from de Groot et al. 2012).
Rp 48.5 m/haP10–P90 Rp 32.6–68.8 m/hagrade D
de Groot, R. et al. 2012. Global estimates of the value of ecosystems and their services in monetary units. Ecosystem Services 1: 50-61 (biome means). Benefits transferred at 0.25-1× the mean (most likely 0.5). Cultivated land: no extra non-market value counted.
CarbonCarbon credits from replanting, after buffer, leakage and uncertainty deductions, at a placeholder price.
Rp 2.6 m/haP10–P90 Rp 1.12–4.81 m/hagrade D
Restoration Studio carbon profiles × (1 − buffer − leakage − uncertainty) × carbon price, minus measurement and issuance costs
Avoided downstream impactsmodel valueAvoided sediment, flooding and water-treatment costs downstream. A model value, not a measurement.
Rp 16.3 m/haP10–P90 Rp 8.73–27.7 m/hagrade D
MODEL value: avoided sediment, flooding and water-treatment costs downstream of restored mining land, Rp/ha/yr. No Indonesian per-hectare study found; ESMI assumption, to be replaced by causal estimates.
RemediationRegrading, soil and water testing, labour, and POCO compost with microbes where metals are expected.
Rp 32.5 m/haP10–P90 Rp 27.9–37.9 m/hagrade D
Restoration Studio remediation costs: regrading, XRF and lab sampling, labour, POCO + microbes on the share of land with metals
PlantingSeedlings, compost and planting for the plan mix.
Rp 27.6 m/haP10–P90 Rp 22.9–33.6 m/hagrade D
Restoration Studio option costs (planting, seedlings, compost) of the plan mix
Restoration costsRemediation and planting, plus a contingency.
Rp 65 m/haP10–P90 Rp 55.3–76.8 m/hagrade D
Remediation + planting + contingency
Reclamation duty dischargedThe reclamation duty that restoration would discharge. It is a transfer between parties (who pays): shown, not added to the gain.
Rp 177 m/haP10–P90 Rp 114–257 m/hagrade D
Reclamation standard (Kepmen ESDM 344.K/MB.01/MEM.B/2025, IKK-adjusted): the duty restoration discharges. A TRANSFER (who pays), not added to the gain
Net gainmodel valueGain in land value + ecosystem services + carbon + avoided downstream impacts − restoration costs.
Rp 68.4 m/haP10–P90 Rp 31.6–107 m/hagrade D
Net gain = (value after − today's use value) + ecosystem services + carbon + avoided downstream impacts − costs; present value, model estimate, not an appraisal
Mining footprint: Cheng et al. 2026, Maus et al. 2022 (v2), ESMI digitised; counted once · 31 Dec 2022
Values in Rp (ecosystem services) are a data gap until ESVD valuation is connected. · il.v1
What is the EAI (Environmental Affected Index) of this area, and which component weighs most?
Answered
9.7EAI
EAI 9.7 (low). Heaviest component: Permits in protected areas. 56% of components have data.
For Kabupaten Batanghari
Grade C
Sources · Method
Eco Mining Data EAI v1 (Environmental Affected Index; method in STATUS-answers.md)
Mining footprint: Cheng et al. 2026, Maus et al. 2022 (v2), ESMI digitised; counted once · Maus, V. et al. (2022) Scientific Data 9, 289; Cheng et al. (2026) · CC BY 4.0 / CC BY-SA 4.0
Method: Sentinel-5P tropospheric NO2 column enhancement over a local background, mean of the newest 12 months (strongest of the regency mean and its industrial zones). Satellite column density, not a ground concentration and not a health exposure value
Partial answer: ISPU stations in the region; Sentinel-5P, CAMS and NEA are not connected yet.
Missing: CAMS (Copernicus atmosfeer: analyses en voorspellingen) · NEA Singapore (PSI/PM2.5 per uur, data.gov.sg) + ASMC (ASEAN-nevelkaarten)
Where do most new alerts come from this week/month?
Answered
0alerts
0 confirmed alerts in the last 30 days (0 in the last 7 days). Newest alert: 24 Jun 2026.
For Kabupaten Batanghari
Grade C
Data as of 24 Jun 2026 outdated
Details to verified organisations
Sources · Method · Partial answer
Eco Watch alerts (Eco Mining Data; confirmed after four-eyes review)
Method: Eco Watch alerts in the last 7 / 30 days; public count = confirmed only, unconfirmed signals counted for T3 in region scope
Partial answer: Eco Watch alerts (confirmed in public; unconfirmed signals counted for verified organisations only). RADD/GLAD and own detection are not live yet.
Social: Conflicts, displacement, safety (accidents, collapses at PETI sites)
Answered
2events
2 reported conflicts, accidents or disasters in 24 months (0 confirmed). Most frequent: Spill / tailings leak.
For Kabupaten Batanghari
Grade D
Data as of 18 Oct 2026
Sources · Method · Partial answer
Eco Mining Data news engine (headlines as published; events extracted, counted) · 18 Oct 2026
Method: Media-reported conflicts, accidents, disasters and spills (24 months); counted, mostly unconfirmed, no organisation named
Partial answer: Media-reported conflicts, accidents and disasters (counted, mostly unconfirmed, no organisations named); police and field data not yet.
Missing: GDELT (wereldwijd nieuws, geotagged) · Field reports
Regency mean compared with the background of its island group
Grade C
Satellite column density, not air at ground level. Sentinel-5P (TROPOMI) measures the amount of NO₂ and SO₂ in the whole air column. The figure is the ENHANCEMENT over the surrounding background, not a ground-level concentration and not a health exposure value. The zone is named by place; this is not a finding about any company.
0.19× 10¹⁵ molecules/cm² NO₂ above backgroundmean of the newest 12 months (to 2026-09), ± 0.01 (1 SE, statistical)No significant trend (0.00 per year, p = 0.917, 9 years)
Monthly NO₂ enhancement (thin), 12-month running mean (bold), ± 1 SE band; 0 = background
SO₂ for this place is not available yet (sampled orbits still being processed).
The ±SE is statistical only; the retrieval (air-mass factor) uncertainty is 25–50 % for NO₂ and larger for small SO₂ columns. No wind data is used yet: the background ring covers all directions. Sources: KNMI TEMIS monthly TROPOMI NO₂ v1.0; Copernicus Sentinel-5P OFFL L2 SO₂ via Microsoft Planetary Computer.
