Wetland Management Session

Transcription

1 2 nd International Seminar UNFCCC On Cutting Workshop Edge on Science & Technologies technical towards and Food, scientific Environment aspects of ecosystems and Health with - Focus: high-carbon Civil reservoirs Society not covered Göttingen, by other Germany, agenda items 2-4 September under the Convention October 2013 United Nations Campus Altes Abgeordneten-Hochhaus, Bonn, Germany Wetland Management Session Challenging on Eco-System of Carbon- Water-Soil in Tropical Peatland and Wet/lowland in South East Asia - Innovation of Concept-Technology-Land Management- Session 1. Current Scientific and Technical Knowledge Mitsuru Osaki 1, Takashi Hirano 1, Kazuyo Hirose 2, Hidenori Takahashi 3, Yukihiro Takahashi 4, Gen Inoue 5, Muhammad Evri 6, and Bambang Setiadi 6 Estimating Carbon Fluxes and Stocks through integrated monitoring in Tropical Peatland 1 Research Faculty of Agriculture, Hokkaido University, Japan 2 Earth Remote Sensing Division, Japan Space Systems, Japan 3 NPO HKD Institute Hydro-climate, Japan 4 Graduate School of Science, Hokkaido University, Japan 5 Research Member of SATREPS, Japan Mitsuru Osaki 6 Agency for the Assessment and Application of Technology, Research Faculty of Agriculture, Hokkaido University, Japan Indonesia

2 Introduction Photo from Erianto Indra Putra (UNPAR)

3 Amount of Carbon in Tropical Peat (GtC (%)) Others 1.51Gt (2.0) West Papua 10.3Gt (23.0) Indonesia 44.5Gt (100) Sumatra 18.3Gt (41.1) Kalimantan 15.1Gt (33.8) Other tropical area 25.7Gt (30.7) Other Southeast Asia 13.6Gt (16.2) Tropical area 83.8Gt (100) Indonesia 44.5Gt (53.1) (From Maria Strack ed., 2008: Peatlands and Climate Change. International Peat Society, 223pp.)

4 Total amount of CO 2 emission

5 Integrated Monitoring-Sensing- Modeling (MSM) system Photo from Erianto Indra Putra (UNPAR)

6 Main Project Sites Monitoring was started from 1997 Central Kalimantan, Indonesia Peatland area in Mega Rice Project site CO 2 observation towers at UDF:(Un-drained Peat) DF:(Drained Peat ) BC:(Burnet Peat) UD F DF BC Various Study Topics: GHG Flux (CO 2, CH 4, N 2 O) measuring Fire Detection and Protection Water Table Monitoring and Management Peatland Ecology Soluble Carbon Monitoring Peatland Subsidence Monitoring etc.

7 Key elements for integrated Monitoring-Sensing- Modeling (MSM) system of Carbon in peatland (1) CO 2 Flux & Concentration CO 2 Flux (2) Wildfire detection & Hotspot 1. Atmospheric Elements (3) Forest degradation & Species mapping (4) Deforestation & Forest biomass change 2. Above Ground Elements (Forest Biomass) (5) Water level, & Soil moisture 3.Water Elements Peat (Carbon) Drilling (6) Peat thickness & Peat dome detection (7)Peat subsidence Water Gauge (8)Water soluble organic carbon 4. Below Ground Elements (Below Ground Carbon Stock)

8 Satellite Micro-Satellite & LCTF* 4 GOSAT (1) Terra & Aqua MODIS (2) Landsat, SPOT, TerraSAR, AVNIR-2, VHR* 2 Sensors (3), (4) ASTER, Hisui (3), (4), (8) PALSAR, AMSR-E (4), (5), (6), (7) Airborne /***UAV UAV* 3 (1), (3) LiDAR (4), (6), (7) Ground DGPS(7) Lateral CO 2 Flux Vertical CO 2 Flux Tower(1) Chamber(1) (1) CO 2 Flux & Concentration (2) Wildfire detection & Hotspot (3) Forest degradation & Species mapping FES-C* 1 (1) (4) Deforestation & Forest biomass change (5)Water level, & Soil moisture DGPS(7) Drilling(6) *1:FES-C : Fiber Etalon Solar measurement of CO 2 *2:VHR : Very High Resolution Remote Sensing Data *3:UAV: Unmanned Aerial Vehicle *4:LCTF: Liquid Crystal Tunable Filter Red: Instrument Black: Target (6)Peat dome detection & Peat thickness (7)Peat subsidence Key Elements of Tropical Peatland MSM System Water Gauge(5) (8)Water soluble organic carbon

9 Integrated Monitoring-Sensing- Modeling (MSM) system: Carbon Stock Photo from Erianto Indra Putra (UNPAR)

10 Peat Thickness Estimation (Shimada Model) Vegetation (Physiology) Adjust Response Phenology Classification (cf. Lugo et al. 1990) Hydroperiod Shallow peat layer Mineral Soil WET DRY Greater GWL fluctuation Hypothesis Temporarily High flood condition Deep peat layer Mineral Soil WET DRY Moderate GWL fluctuation Permanently Wet condition Seasonal fluctuation of GWL is different; moderate at deep peat, while greater at shallow peat. This might reflect the vegetation physiologically.

11 Idea of Peat Depth Classification In Tropical Peat Swamp Forest, type of forest stand and its phenology are corresponded to Peat Depth, in terms of seasonal groundwater level fluctuations. Its difference produce spatial trends of plant activity in each season. To detect these, Supervised classification were conducted using multi-temporal satellite scene with Peat Depth Database as training data. Index of Plant Activity: NDVI Target Period : Early 90 s Relatively Undisturbed Condition (Before Mega Rice Project) Marginal Mixed Swamp Forest Elevation (m) NDVI Low Pole Forest Tall Interior Forest Distance from river (km) NIR NIR Multi-temporal satellite scene (NDVI) were assembled Red Red A month NDVI B month NDVI C month NDVI D month NDVI. Peat Surface. Mineral ground Shepherd et al. (1997) with partial modification

12 N Peat thickness (m) 0 2 Estimated Map of Peat Thickness Root Mean Square Error (RMSE) = 1.64 m 4 Ground-truth peat thickness (m) Distribution Map of C-density (Shimada et al. 2001) N C density (kg m -2 ) Mha C-pool = 4.2 Gt 300 < 1.4 Gt Mha -1

13 Classified map Classification were conducted within the area below 1) Estimated Swamp Forest extent built from Landsat image (1994) and SRTM DEM 2) PalangkaRaya & Pulang Pisau Regency where include core research area of SATREPS We are still trying to collect peat drilling data with depth infomation to rebuild the map

14 Integrated Monitoring-Sensing- Modeling (MSM) system: Carbon Flux by Oxidation (directly) Photo from Erianto Indra Putra (UNPAR)

15 Montly-mena NEE (gc m -2 d -1 ) Seasonal variation in net CO 2 exchange (NEE) NEE Year NEE = RE - GPP UF DF BD Large increases were found in the dry seasons of 2002, 2004 and 2006, El Niño years, because of shading by dense smoke and the enhancement of oxidative peat decomposition due to low GWL.

16 NEE (gc m -2 y -1 ) Annual NEE vs. annually mean GWL UF DF DB GWL (m) Hirano et al., 2012 A negative linear relationship for each site Enhancement of oxidative peat decomposition under low GWL Slope: UF > DF > DB Undisturbed peatland is more sensitive. Annually mean GWL is a robust indicator to assess annual CO 2 balance.

17 Oxidative peat decomposition vs. GWL in burnt site Burnt peatland With 6 automated chambers Fires Little vegetation From 2004 to 2005 Heterotrophic respiration (oxidative peat decomposition) Peat decomposition (RS) From a simple relationship, GWL lowering by 0.1 m Additional peat decomposition of 89 gc m -2 y -1 Hirano et al., 2013 (GCB)

18 Integrated Monitoring-Sensing- Modeling (MSM) system: Carbon Flux by Oxidation (indirectly) Photo from Erianto Indra Putra (UNPAR)

19

20 Factors affecting on the level of ground surface in tropical peatland (1) Weight Effect? Biomass Biomass: 0.1 t m -2 Own weight of peat: t m -2 (1 m) Buoyancy of peat: 0 t m -2 Peat Groundwater Water/Air Gap Woody matrix Mineral soil

21 Factors affecting on the level of ground surface in tropical peatland (2) Bulk Density? Woody matrix and water/air gaps (Yonebayashi et al, 1995)

22 Factors affecting on the level of ground surface in tropical peatland (3) How separate Decomposition and Compaction? Biomass (1 m) Decomposition and Compaction Peat Groundwater Mineral soil

23 GEOSCANNER ASSESSMENT Sumber arus A M N B Sumber arus Permukaan bumi equipotensial equipotensial Arah arus Arah arus

24 Subsidence measurement by use of a laser distance meter The laser distance meter is fixed on the pole which is inserted into the soil until it reaches to a clay soil. The target of observation is a plastic pole put on the ground removing the fresh litter on surface. The change of the distance is recorded together with the underground water level. This change corresponds to the peat soil layer thickness change above clay layer. The precision of distance meter is 50 μm.

25 Subsidence (mm) Ground water table (cm) Peat surface level (mm) Peatland Subsidence Monitoring Peat is shrinking and swelling which is strongly affected by ground water table. It results in uplift and subsidence which are measured by five methods with different precision (Fig. A). Laser distance meter method enables to monitor the subsidence precisely and reliably (Fig. B) Av. precision (5) Laser distance meter (1) (2) (3) (4) (5) Fig. A: Precisions of five subsidence methods; (1) Light Detection and Ranging (LiDAR), (2) Interferometric SAR(InSAR), (3) Leveling pole, (4) differential GPS, (5) Laser distance meter /17 4/6 4/26 5/16 6/5 6/25 Fig. B Relationship between ground water level (blue) and peat surface level (red) Subsidence CO2 emission Presented by Kawasaki (2013) Drainage Peat (Carbon)

26 Integrated Monitoring-Sensing- Modeling (MSM) system: Carbon Flux by Fire (directly) Photo from Erianto Indra Putra (UNPAR)

27 How to evaluate the CO 2 flux from wild fire? Out-flow In-flow Flux = C(h,x) v(h,x) dx dh v: Wind velocity at (h,x) C: Concentration at (h,x) Flux is evaluated from concentration of CO 2, wind speed and the plume distribution

28 CO2 concentration (ppm) S2.5 S3.0 CO2 emission (Mt in ) FRP NDVI GFED This work 42± ± ± (in 2011)

29 Integrated Monitoring-Sensing- Modeling (MSM) system: Carbon Flux by Modeling Photo from Erianto Indra Putra (UNPAR)

30 Lowest groundwater (cm) Depth from ground surface 0-50 Takahashi Model The lowest GWL in dray season and Peat Fire Index Summary 1. Peat boundary selection 2. GWT by Takeuchi model 3. CO2 emission by Hirano model 4. Fire occurrence by Takahashi model -100 y = x - 52 R² = Peat Fire Index correlation Fire occurrence (Takahashi) (Takahashi)

31 NEE (gc m -2 y -1 ) Hirano Model Annual NEE vs. Annually Mean GWL UF DF DB GWL (m) UF NEE=-2376GWL-151 DF NEE=-1609GWL-510 DB NEE=-789GWL-378 NEE: Net Ecosystem CO2 Exchange Hirano et al.(2012), GCB Takeuchi Model GWT estimation by Remote Sensing Data GSMaP MTSAT In-situ ground water Precipitation table Land Surface Temp. Drought Index Ground Water Table Takeuchi, Hirano, Anggraini and Roswintiarti (2010) AMSR

32 Water Table Mapping Satellite Sensing Modeling Algorism Tuning By Wataru Takeuchi, University of Tokyo, Japan Water Table Mapping Input Output Coefficiency between Water Table Level and 1) CO2 emission by Oxidation 2) CO2 emission by Fire Factors Mapping of 1) CO2 emission by Oxidation 2) CO2 emission by Fire Factors

33 CO 2 balance (NEE, tc m -2 month -1 ) of peatland in Central Kalimantan in 2011 By Park and Takeuchi, University of Tokyo

34 The Ecosystem model developed by NIES(Dr. Ito) Objectives Vegetation Integrated SImulator for Trace gases (Developed in NIES & FRCGC-JAMSTEC) Atmosphere-ecosystem biogeochemical interactions Especially, major greenhouse gases (CO 2, CH 4, and N 2 O) budget Assessment of climatic impacts and biotic feedbacks Point-global, daily-monthly - CO 2 : photosynthesis & respiration - CH 4 : production & oxidation - N 2 O: nitrification & denitrification - LUC emission: cropland conversion - Fire emission: CO 2, CO, BC, etc. - BVOC emission: isoprene etc. - Others: N 2, NO, NH 3, erosion Carbon-cycle (Sim-CYCLE-based) Nitrogen-cycle

35 Satellite GOSAT IBUKI Senescing: CO2 Column averaged dry air mole fraction distribution of carbon dioxide for the month of September, 2009, obtained from IBUKI observation data (unvalidated) By JAXA Simulator: SimCycle-Visit for East Asia

36 Improved scheme for soil respiration of tropical peat forest Water table (WT) Precipitation (PT) Evapotranspiration (ET) Soil temperature (Ts) Tank model Soil water content (SWC) Soil water potential (SWP) Improved soil respiration function Soil respiration

37 Integrated Monitoring-Sensing- Modeling (MSM) system: Conclusion Photo from Erianto Indra Putra (UNPAR)

38 Top-down satellite airplane inverse model Satellite GOSAT IBUKI Senescing: CO2 Carbon-Water Simulation Integrated, practical carbon budget map Column averaged dry air mole fraction distribution of carbon dioxide for the month of September, 2009, obtained from IBUKI observation data (unvalidated) By JAXA Simulator: SimCycle-Visit for East Asia Bottom-up field survey flux obs. process model Subsidence Model Carbon Emission by Fire Carbon Loss through Water Carbon Emission by Microorganisms Degradation Tree Growth/Mortality Pest subsidence

39 Attention to environmental catastrophe! Thanks for your attention!