## New Research In

### Physical Sciences

### Social Sciences

#### Featured Portals

#### Articles by Topic

### Biological Sciences

#### Featured Portals

#### Articles by Topic

- Agricultural Sciences
- Anthropology
- Applied Biological Sciences
- Biochemistry
- Biophysics and Computational Biology
- Cell Biology
- Developmental Biology
- Ecology
- Environmental Sciences
- Evolution
- Genetics
- Immunology and Inflammation
- Medical Sciences
- Microbiology
- Neuroscience
- Pharmacology
- Physiology
- Plant Biology
- Population Biology
- Psychological and Cognitive Sciences
- Sustainability Science
- Systems Biology

# How temporal patterns in rainfall determine the geomorphology and carbon fluxes of tropical peatlands

Edited by Donald E. Canfield, Institute of Biology and Nordic Center for Earth Evolution, University of Southern Denmark, Odense M., Denmark, and approved May 5, 2017 (received for review February 6, 2017)

## Significance

A dataset from one of the last protected tropical peat swamps in Southeast Asia reveals how fluctuations in rainfall on yearly and shorter timescales affect the growth and subsidence of tropical peatlands over thousands of years. The pattern of rainfall and the permeability of the peat together determine a particular curvature of the peat surface that defines the amount of naturally sequestered carbon stored in the peatland over time. This principle can be used to calculate the long-term carbon dioxide emissions driven by changes in climate and tropical peatland drainage. The results suggest that greater seasonality projected by climate models could lead to carbon dioxide emissions, instead of sequestration, from otherwise undisturbed peat swamps.

## Abstract

Tropical peatlands now emit hundreds of megatons of carbon dioxide per year because of human disruption of the feedbacks that link peat accumulation and groundwater hydrology. However, no quantitative theory has existed for how patterns of carbon storage and release accompanying growth and subsidence of tropical peatlands are affected by climate and disturbance. Using comprehensive data from a pristine peatland in Brunei Darussalam, we show how rainfall and groundwater flow determine a shape parameter (the Laplacian of the peat surface elevation) that specifies, under a given rainfall regime, the ultimate, stable morphology, and hence carbon storage, of a tropical peatland within a network of rivers or canals. We find that peatlands reach their ultimate shape first at the edges of peat domes where they are bounded by rivers, so that the rate of carbon uptake accompanying their growth is proportional to the area of the still-growing dome interior. We use this model to study how tropical peatland carbon storage and fluxes are controlled by changes in climate, sea level, and drainage networks. We find that fluctuations in net precipitation on timescales from hours to years can reduce long-term peat accumulation. Our mathematical and numerical models can be used to predict long-term effects of changes in temporal rainfall patterns and drainage networks on tropical peatland geomorphology and carbon storage.

- tropical peatlands
- peatland geomorphology
- peatland hydrology
- peatland carbon storage
- climate-carbon cycle feedbacks

Tropical peatlands store gigatons of carbon in peat domes, gently mounded land forms kilometers across and 10 or more meters high (1). The carbon stored as peat in these domes has been sequestered by photosynthesis of peat swamp trees (2) and preserved for thousands of years by waterlogging, which suppresses decomposition. Human disturbance of tropical peatlands by fire and drainage for agriculture is now causing reemission of that carbon at rates of hundreds of megatons per year (2⇓⇓–5): Emissions from Southeast Asian peatlands alone are equivalent to about 2% of global fossil fuel emissions or 20% of global land use and land cover change emissions (6, 7). Because peat is mostly organic carbon, a description of the growth and subsidence of tropical peatlands also quantifies fluxes of carbon dioxide (1, 4, 8). Evidence from a range of studies establishes that accumulation and loss of tropical peat are controlled by water table dynamics (4, 9). When the water table is low, aerobic decomposition occurs, releasing carbon dioxide; when the water table is high, aerobic decomposition is inhibited by lack of oxygen, production of peat exceeds its decay, and peat accumulates. In this way, the rate of peat accumulation is determined by the fraction of time that peat is exposed by a low water table (Fig. 1).

The water table rises and falls in a peatland according to the balance between rainfall, evapotranspiration, and groundwater flow. Water flows downslope toward the edge of each peat dome, where it is bounded by rivers. This flow occurs at a rate limited by the hydraulic transmissivity of the peat—the efficiency with which it conducts lateral flow—and follows the gradient in the water table. The gradient in the water table is slightly steeper near dome boundaries where the flow of water is faster. A steeper gradient near boundaries implies a domed shape in the water table, or groundwater mound, corresponding to the domed shape of the peat surface. The doming of the peat surface is very subtle: Gradients are about 1 m/km (1). Nonetheless, it is the dome’s gentle curvature that accounts for the carbon storage within the drainage boundary.

Once the peatland surface is sufficiently domed, water is shed so rapidly that no more organic matter can be waterlogged within the confines of the drainage network, and peat accumulation stops (10). This maximally domed shape sets a limit on how much carbon a peat dome can sequester and preserve under a given rainfall regime (11). If the peat dome is flatter than its stable shape for the current climate, it will sequester carbon and grow; if it is more domed than its stable shape, it will release carbon and subside as peat decomposes. [In the tropical peat literature, “subsidence” is used for a decline in the peat surface elevation, regardless of mechanism (5).] The volume of this stable shape times the average carbon density of the peat defines a capacity for storage of carbon as peat within the drainage boundary.

If we can predict the stable shapes of peat domes and how they evolve over time in a given climate, we can determine how peatland carbon storage capacity and carbon fluxes are affected by changes in rainfall regime, drainage network, and sea level. However, when predicting the stable shapes of peat domes and their evolution toward these shapes, there are two complicating factors: (*i*) The boundaries imposed by drainage networks have complex shapes and (*ii*) rainfall is intermittent and variable. The water table rises during rainstorms and falls during dry periods, even when the peat surface is stable. These fluctuations in the water table seem to be important because it is widely believed that seasonality of rainfall affects tropical peat accumulation (12, 13). But how should we take these fluctuations into account to predict the slow development and stable shapes of peat domes? Understanding the global impact of changes in rainfall amount and variability, drainage networks, and sea level on tropical peatland carbon storage and fluxes requires a theory that can accommodate the complicated drainage networks and intermittent rainfall of the real world.

Ingram (10) made the first prediction of the limiting shape of a temperate peat dome imposed by the balance between rainfall and groundwater flow. Assuming constant rainfall, he computed the steady-state shape of a peat dome with uniform permeability between parallel rivers. Clymo (14) later developed a simple dynamic model for accumulation of peat at a single point in the landscape. Clymo’s model assumed that the thickness of peat above the water table would not change and focused on anaerobic decomposition in deeper waterlogged peat. Hilbert et al. (15) later built on Clymo’s model to allow a varying thickness of peat above the water table via a simple water balance whereby drainage increases linearly with peat surface elevation. Hilbert’s model inspired a series of increasingly sophisticated models for vegetation dynamics and peat accumulation at a point. The most recent of these point models computes water table depth from monthly rainfall, using a site-specific model (16). Meanwhile, numerical models have been used to simulate peat accumulation under constant rainfall (17, 18). Although these subsequent works simulate the dynamics of peat production and decomposition in increasing detail, a strength of Ingram’s model was that it provided quantitative intuition for how peat dome morphology depends on peat hydrologic properties and average rainfall. Could a principle like Ingram’s exist that describes peatland dynamics as well as statics and remains applicable with realistic drainage networks and rainfall regimes?

We established a field site in one of the last pristine peat swamp forests in Southeast Asia and then used measurements from this site to develop a mathematical model for the geomorphic evolution of tropical peatlands that is simpler, yet more general than Ingram’s model for high-latitude peatlands. Our model makes it possible to predict effects of changes in rainfall regime and drainage networks on carbon storage and fluxes in tropical peatlands. The model predicted, perhaps surprisingly, that surface peat would be older near dome margins. We tested these predictions by radiocarbon dating core samples and comparing the age of each sample to the simulated age at its location and depth. Finally, we explored the future of tropical peatlands under climate projections by simulating the geomorphic evolution of an idealized peat dome under projected changes in rainfall patterns and drainage.

## Methods

### Field Measurements.

We established a field site in a pristine peat forest in Brunei Darussalam (Borneo) to study a peat dome where current processes affecting peat accumulation are essentially similar to those during its long-term development (Fig. 2). At the site, we installed 5 piezometers along a 2.5-km trail, 12 piezometers along a 180-m transect, and 3 throughfall gauges. We completed a total station survey of peat surface elevation along the transect to characterize peat surface microtopography. To characterize large-scale peatland morphology, we also obtained LiDAR data for the entire study area. To study peat dome development, we collected nine peat cores from which we obtained 35 radiocarbon dates. To test whether our undisturbed site behaved similarly to sites studied by other groups, we installed four soil respiration chambers and a piezometer at a nearby logged but undrained site.

### Morphology vs. Microtopography.

Superimposed on the gross morphology of a peat dome is a fine microtopography of meter-scale depressions, or hollows, separated by higher areas, or hummocks (19, 20). The hummocks consist of partly decomposed logs, branches, and leaves lodged among living buttresses, stilt roots, pneumatophores, and giant rhizomes. Whereas the microtopography in high-latitude peat bogs may have regular and oriented patterns (21), surveys by Lampela et al. (20) in a tropical peat swamp in Central Kalimantan showed no orientation or regularity. Similarly, our microtopography survey and other observations revealed no regular patterns or channels in peat dome microtopography.

In describing the evolution of peat dome morphology, we want to capture the effects of the hummock-and-hollow microtopography without explicitly simulating its details. Measurements from the 12 piezometers along our microtopography transect showed that the water table is relatively smooth, even though the peat surface is highly irregular on a spatial scale of centimeters to meters (Fig. 3). We therefore represent the peat surface by a reference surface *SI Methods*).

### Groundwater Flow.

We model the dynamics of the water table

At high water tables, hollows become flooded from saturation of the peat below, forming small pools. These pools are not connected into channels (20) and therefore do not allow open-channel flow on a large scale in the peatland. Instead, flow through the peatland is limited by flow through the porous matrix of the hummocks between these isolated pools. We apply Boussinesq’s equation at scales much larger than hummocks and hollows (tens of meters) and refer to the flow of water through the peatland as “groundwater flow” even though some flow occurs above the local peat surface, in hollows, during wet periods. Boussinesq’s equation requires only that lateral flow is proportional to the head gradient, which is the case if the overall flow is limited by laminar flow through hummocks. We never observed ephemeral channels connecting hollows within the peatland in our 6 y at the site. In addition, if flow were nonlaminar, we would expect different local flow behavior at the same water table height in areas with different water table gradients, but instead water table behavior is uniform (*Results and Discussion*).

### Local Carbon Balance.

A broad range of studies demonstrates that the thickness of peat exposed above the water table determines the rate of peat accumulation or loss (4, 22). Like others (4, 22), we modeled the dynamics of peat accumulation or loss

Several other studies have shown a leveling off of soil CO_{2} efflux at very low water tables (25, 26), and it is also likely that very high water tables ultimately limit net carbon uptake by trees (primary production) (16). However, including these effects did not affect simulations because these extreme water table heights and depths were neither observed at our site nor predicted by simulations of our site. We also did not include anaerobic decomposition below the water table because analyses of peat cores from tropical sites in Asia (2), including our site (27), do not show detectable loss of waterlogged peat from anaerobic decomposition.

### Numerical Simulations.

We built a numerical model of waterlogging and peat accumulation based on Eqs. **1** and **2** to simulate peat dome geomorphogenesis and carbon fluxes. These two equations are coupled by the water table elevation *i*) a specific yield function *ii*) a transmissivity function *iii*) a rate of peat production *iv*) a decomposition rate constant *SI Expanded Description of Peat Dome Simulation*).

We determined the specific yield and transmissivity functions *Results and Discussion*). We then fitted the parameters for peat accumulation

### Limitations of Modeling Approach.

Our goal was to build the simplest model that can make reasonable quantitative predictions of tropical peat dome dynamics. In most Southeast Asian peatland complexes, every area between rivers is occupied by a peat dome, so it is not apparent how any peat dome could now expand to fill a larger area. However, domes tend to be larger in older peatlands, suggesting a long-term process of dome coalescence. We did not attempt to model these long-term changes in river networks. We also did not consider changes in hydraulic conductivity near the surface caused by compaction or changes in microtopography under agriculture.

## SI Methods

### Field Data Collection.

Field data were collected in the Ulu Mendaram Conservation Area of the Belait District in Brunei Darussalam (4.359863° N, 114.352252° E; Fig. 2). The site is described in more detail in Dommain et al. (27). Average rainfall at the nearby Seria and Kuala Belait weather stations is 2,880 mm/y (1947–2004). The area has never been logged in recorded history, although commonly domesticated tree species planted along the river suggest that there were settlements there some decades ago. The peat dome shows a typical northwest Borneo peat swamp catena (1), with mixed swamp forest at the river’s edge, followed by an Alan Batu community (scattered, tall *Shorea albida* in upper canopy, large gaps mostly dominated by *Pandanus andersonii*) and then an Alan Bunga community farthest from the river (closed upper canopy of *S. albida* 50 m tall). The forest floor is a dense tangle of buttresses, *Pandanus* rhizomes, and woody debris, with no clearly defined channels for the flow of water (Fig. 3). Data are available from the Dryad Digital Repository (dx.doi.org/10.5061/dryad.18q5n).

#### Soil respiration.

Soil respiration measurements were made with an automated dynamic chamber system (LI-8100 with LI-8150 multiplexer; LI-COR Biosciences), in which carbon dioxide concentrations were measured after periodic closure of four chambers connected to an infrared gas analyzer via a gas multiplexer. Each chamber was closed for 6 min and a flux computed every hour, using instrument software (LI-8100 software version 2.0.0). Chambers were supported 30 cm above the peat by four 2-inch PVC pipes, 1 m long, to avoid submergence of the chambers. Although temperature affected individual flux measurements around the diurnal cycle, the effect of temperature on daily mean fluxes was negligible. Water table height was simultaneously monitored using a logging pressure transducer as described below (piezometers).

#### Microtopography transect.

We used a total station [TC-GTS-105N (60536) 5“ Totalstation; B. L. Makepeace, Inc.] to survey peat microtopography and 12 piezometers along a 180-m transect at the field site in August 2012 (Fig. 2 *D* and *E*). We refer to the piezometers in this transect as “transect piezometers” to distinguish them from the 5 “trail piezometers” along the 2.5-km trail. We located the transect survey points in geographic space by rotating the points to match the compass orientation between the first two piezometers and then shifting the rotated points so that the position of the first piezometer matched its GPS coordinates (Fig. 2*E*).

To obtain the elevation of the survey points above mean sea level, we matched the survey data to the LiDAR data. The LiDAR digital terrain model (DTM) was defined by finding local minima in last-return elevations on a 20-m × 20-m grid. Therefore, we found local minima in surveyed peat surface points on the same grid and then aligned the (arbitrary) vertical coordinate of the minimal survey points and the DTM elevation in the corresponding grid cell (Fig. 3*B*).

We then defined a reference land surface *B* and *C*). We defined the land surface *B*). By this definition, the land surface

We next used the transect hydrologic and microtopographic data to define the land surface at each trail piezometer. Our definition of the land surface *C* and 5*A* ). Therefore, to find the land surface elevation for each trail piezometer, we first found the land surface for this nearest piezometer by matching the data on the overlapping time interval. Because the behavior of the water table was so similar in all trail piezometers except the bog plain piezometer, for plotting in Fig. 5 we simply matched the water table height

#### LiDAR data and topography.

LiDAR data were obtained from the Brunei Survey Department from a 2009 aerial mission. We constructed a DTM from the LiDAR data by taking the minimum last-return elevation within 20-m *C*). Although the groundwater divide is on the Malaysian side of the border where LiDAR data are unavailable, the tapered shape of the flow tube ensures that the simulations are insensitive to the precise location of the divide (Fig. 6).

We sampled the magnitude of the gradient along each contour, using a spatial database (SpatiaLite 4.2.0; www.gaia-gis.it/gaia-sins/). By the divergence theorem, the average Laplacian of surface elevation within an enclosed region is equal to the normal surface gradient integrated around its boundary divided by the enclosed area. We estimated the Laplacian in the region around the piezometer by linear regression of the integrated normal gradient against the enclosed area in the four piezometers nearest the river, where the slope was approximately uniform (Fig. 6).

#### Peat cores.

Methods for the two peat cores and 13 samples taken for radiocarbon dating with a modified Livingstone corer are available in Dommain et al. (27) and those for the seven peat cores and 22 radiocarbon samples taken with a Russian peat sampler are in Gandois et al. (31). Median calendar ages were estimated from radiocarbon ages, using Calib 7.1.0 with the IntCal13 calibration curve (36, 37). Because cores were taken in hollows, the top of each core corresponds roughly to the reference land surface

#### Piezometers.

Water table height was recorded every 20 min at each of five piezometers along a 2.5-km transect, using logging pressure transducers (Solinst Levelogger Edge 3001; Solinst Canada) (Fig. 2*C*). Transducer measurements were corrected for barometric fluctuations, using a barometer (Solinst Barologger Gold 3001). Each transducer was suspended on a steel cable inside a 2-inch PVC pipe 1.5 m in length installed to 1.4 m depth and screened at 1.3–1.45 m below the top of the piezometer casing. Piezometer locations were determined with a GPS (GPSmap 60csx; Garmin Ltd.).

An additional 12 piezometers were installed along an 180-m transect 1 km from the river (Fig. 2 *D* and *E* and *SI Methods*). These piezometers were screened from 1 m below the peat to 5 cm above the peat and were attached to 2-inch PVC pipes anchored in the underlying clay. Data from these piezometers were recorded from August 17, 2012 through November 3, 2012.

#### Throughfall measurement.

Three siphoning tipping-bucket rain gauges (Texas Electronics TR-525S) were installed along the microtopography transect to record throughfall in the forest understory. Each gauge was mounted on an acrylic plate supported 50 cm above the peat surface on a 2-inch PVC tube pushed into the peat. Large fronds of understory plants (mostly *P. andersonii*) were cleared from above the gauges to reduce the spatial variability in measurements. Tip events were recorded by a switch closure event logger (UA-003-64; Onset Computer Corp.) enclosed in a small PVC case attached to each gauge.

### Calculation of Stable Peat Dome Topography.

We found the stable peat surface Laplacian for a given rainfall regime by a one-dimensional search as follows. The water table in a stable peatland can be simulated efficiently using an integrating ODE solver (SUNDIALS; computation.llnl.gov/projects/sundials). These simulations are fast because the water table behaves the same everywhere, so there is only one hydrologic variable to simulate (Eq. **5**). Initially, we do not know the Laplacian of the stable peat surface **5**) and checking for balance between peat production and decomposition over time (Eq. **3**), successively refining our guess until we have the stable surface Laplacian for the given rainfall regime (Brent’s method).

### Dynamic Model of Peat Dome Development.

We solved local peat accumulation and groundwater flow equations numerically to simulate development of peat domes. We simulated the dynamics of groundwater flow (Eq. **1**) and the dynamics of peat accumulation (Eq. **2**), using a finite volume scheme in one horizontal dimension representing a flow tube on the peat surface (Fig. 2, Fig. S1, and *SI Expanded Description of Peat Dome Simulation*). Simulating tropical peat swamp hydrology is difficult because both transmissivity and specific yield depend strongly on the water table elevation relative to the surface. We avoided the numerical issues that arise from the severe nonlinearity of the flow problem by expressing Boussinesq’s equation in conservation form (38) via a change of variables from the water table elevation to the local water storage (volume per area). The switch to conservation form made it feasible to simulate storm responses over thousands of years with time steps of minutes.

We first split the water table elevation **1**) in terms of the water elevation above the surface **S3**), using a Crank–Nicolson scheme with a no-flow boundary at the groundwater divide and specified surface and water table elevation dynamics at the other boundary (*SI Expanded Description of Peat Dome Simulation*). For each hydrologic time step, multiple internal steps might be taken for stability and accuracy based on a standard stability heuristic (39). After each hydrologic time step, the surface was incremented by straightforward use of Eq. **1**, and the change in water storage below the surface was calculated and deducted from the shallow storage

### Model Calibration.

#### Master recharge and recession curves.

The master recharge and recession curves were computed based on the uniform water table behavior among the four piezometers nearest the river. A related approach, without the assembly of master curves but using the division of water table time series into periods of heavy rain, days without precipitation, and nights without precipitation, has been used previously in other studies of wetland hydrology (28, 29). Under very heavy rain the groundwater divergence term **1**) becomes negligible,*F*) turned on its side. The Laplacian *SI Methods*).

We assembled the master recharge and recession curves by using throughfall data (above) to identify intervals of the time series either with no rain or with heavy rain (rainfall intensity greater than 4 mm/h). Within each interval, we resampled a spline of the head vs. time data to yield time on a uniform grid of water table height. We then found an offset in time or rain depth for each interval to minimize the least-squares difference between overlapping intervals, creating the master recharge and recession curves (*SI Master Recession and Recharge Curve Assembly*). At low water tables, the water table was nearly constant at night and decreased linearly during the day (Fig. 5*F*), so we estimated evapotranspiration as 2.246 mm/d, using the slope of the daytime decrease in water table and specific yield obtained from the master recharge curve.

We estimated parametric functions for the specific yield (cubic spline) and transmissivity (piecewise-linear logarithm of hydraulic conductivity) by simultaneously fitting the recharge and recession curves to model results (Levenberg–Marquardt; www.pesthomepage.org). We simulated recharge [water table height vs. cumulative rainfall depth **S5** for recharge, Eq. **S7** for recession), using an integrating solver (SUNDIALS; computation.llnl.gov/projects/sundials) with the current parameter estimates. Although the data provide no information on how conductivity is distributed in the peat below the lowest excursion of the water table, the dynamics of the water table are always dominated by other factors, either the conductivity of peat higher in the soil profile or evapotranspiration.

#### Calibration for peat surface evolution parameters.

After estimation of hydrologic parameters *A*).

We drove our simulation of the peat dome using a precipitation time series derived from our own throughfall data, meteorological data, and nearby speleothem

### Sensitivity to Environmental and Anthropogenic Change.

We explored the effects of fluctuations in rainfall by using a very simple model for rainfall (Poisson process, exponential depth distribution with two parameters: average storm depth and average interstorm arrival time). Because rain storms are instantaneous in this simple model, we computed the increase in head from each storm by numerical integration of *SI Methods*). We then found the stable Laplacian for each net precipitation time series by root finding as described above (*SI Methods*). As a point of reference for intermittency of rainfall, we calculated the average interstorm arrival time at our site as the average duration of contiguous sequences of 20-min intervals without rain.

In a second set of simulations, we changed the amplitudes of annual and ENSO fluctuations in simulated rainfall time series. We first created a reference rainfall time series by superimposing the 44-y variation in monthly mean rainfall from meteorological data on our throughfall rainfall intensities (as in *SI Methods*, but without superimposing the speleothem signal). We define the amplitudes of the annual and ENSO fluctuations as the amplitudes of sinusoids with arbitrary phases fitted to daily rainfall intensities by least squares. For each desired combination of annual and ENSO amplitudes, we then found the rain time series with the same mean rainfall and desired amplitudes at annual and ENSO periods that was least-squares closest to this reference rainfall time series, while requiring that rainfall intensity was positive or zero and the pattern of storms was the same, using a quadratic programing solver (*SI Setting Annual and ENSO Amplitudes of Rainfall*).

We simulated the effects of environmental and anthropogenic change in the future by imposing different forcing and boundary conditions on simulations starting with an idealized peat ridge between two parallel rivers. We generated the ridge with the simulated precipitation time series used for model calibration and chose the breadth of the ridge so that it would be approximately stable at the end of the time series (2,308 y). The base case for the projections was the reference rainfall time series described above. Current climate projections indicate increased average rainfall, increased seasonality of rainfall, and probably both in the tropics (33). Therefore, we performed simulations with a stronger seasonal signal and scaled the rainfall time series to explore the effects of higher average rainfall. In the first perturbation, we increased the seasonal fluctuations to 3.0 mm/d to match projections for average fluctuations at Brunei’s latitude (4°N) during 2081–2100 (33) (increase of 0.52866 mm/d). In a second projection, we increased mean rainfall by the same amount. The tidal rivers draining coastal tropical peatlands will be directly affected by sea level changes, so in a third perturbation, we increased the boundary water level by 50 cm over an interval of 100 y to mimic the effect of sea level rise (42). [Isostatic subsidence or uplift, such as the postglacial rebound important in many northern peatlands (11), could be accommodated by the model in the same way.] Finally, we simulated the effects on dome shape of drainage to a constant 50 cm below the peat surface, the recommended best practice of the Roundtable for Sustainable Palm Oil (43), to model the effects of conversion to agriculture on dome shape and carbon efflux.

## SI Expanded Description of Peat Dome Simulation

The water table in a tropical peatland can rise by centimeters in a matter of hours in response to intense convective rainstorms, and fluctuations on these short timescales affect the changes in the landscape over millennia (Fig. S4 *D* and *E*). Therefore, simulations required an extremely large number of steps. Two practical consequences of this are that simulations are too large to fit in physical memory on current midrange hardware, and there is a risk of numerical losses in computing cumulants to verify water conservation by the model. To overcome these two issues, we stored forcing data and simulations in hierarchical data format (www.hdfgroup.org/HDF5), periodically prefetching data from the disk to construct interpolants of net precipitation intensity, and computed cumulants using exact arithmetic (Hettinger’s lsum; code.activestate.com/recipes/393090/). The overall simulation driver was implemented in Python (version 2.7.5; www.python.org); rate-limiting calculations were implemented in Cython (version 0.21.1; cython.org).

### Flow Line Coordinate System.

Simulations of groundwater flow and peat accumulation used a flow line coordinate system. Flow lines, contours, and elevation on the peat dome form an approximately orthogonal coordinate system: Streamlines and contours are always orthogonal, and elevation is nearly orthogonal to the plane of the surface because gradients are very small. In general, in orthogonal curvilinear coordinates with three coordinates

Along the reference streamline, the streamline arc length parameter is 1 by definition, and we approximated the contour arc length parameter **S2**) and surface elevation

### Discretization of Groundwater Flow Equations.

We discretized the groundwater flow problem in flow tube coordinates (Eq. **S13**), using a standard one-dimensional finite-volume scheme (39), with a Neumann boundary condition at the origin (zero gradient at groundwater divide) and a Dirichlet condition at the drainage boundary. We use **S13**)

We approximate the gradients with finite differences and find fluxes through the cell faces. Geometry is handled in much the same way in both cases. The shallow fluxes are approximated as

After integrating with respect to time, the discretized problem for interior points

We chose the time step conservatively for stability by requiring that the diagonals of the explicit coefficient matrix for shallow fluxes

## Results and Discussion

### Carbon Storage Capacity of Tropical Peatlands.

#### Local water balance is dominated by flows near the surface.

Eighteen months of data on water table height in five piezometers along a 2.5-km transect (Fig. 5) show two distinctive features of water table behavior in tropical peatlands. First, when the water table is high, it falls very rapidly; and second, the water table height relative to the land surface remains approximately uniform in all piezometers as the water table rises and falls, as observed elsewhere by Hooijer (28). In what follows, we use “water table height” *E* and *F*). During heavy rain, the effects of evapotranspiration and outward flow are negligible, and the rainfall intensity vs. rate of increase in water table height gives the specific yield. Between rain events, the water table declines because of evapotranspiration and the divergence of groundwater flow.

Transmissivity *D*). Therefore, we can obtain the divergence of groundwater flow from the declining water table during dry intervals after accounting for evapotranspiration (refs. 28, 29; further details are in *SI Methods*). We find that transmissivity increases exponentially at high water tables, when water rises into hollows and flows through hummocks, but decreases dramatically at low water tables when water flows through fine pores in the peat matrix (Fig. 5*C*). Very high permeability near the peat surface is consistent with our observations of more void space higher in the peat profile and also with recent data from other tropical peatlands (30). The water table curves (Fig. 5 *E* and *F*) indicate that the near-surface permeability is so great that the total thickness of deeper peat is unimportant for groundwater flow. Therefore, transmissivity is approximately independent of peat depth and depends only on the water table height

#### Morphology of peat surface explains uniform water table behavior.

According to Boussinesq’s equation, uniform transmissivity is not, by itself, enough to explain the uniform fluctuation of the water table. Even in hydrologic systems where hydraulic properties are uniform, the water table can behave differently at different locations because of topography. For example, in most hydrologic systems a rainstorm drives a different water table response at a topographic divide than it does near where groundwater discharges to a river.

To understand the uniform water table behavior in peatlands, we refer back to Boussinesq’s equation (Eq. **1**). If both the specific yield *C–E*). (The “Laplacian of the peat surface” **1**) in terms of the water table height relative to the land surface (

The peat surface Laplacian describes the curvature of the peat surface: It is equal to the sum of the second derivatives of the surface elevation in two perpendicular horizontal directions (

Maps of the peat surface Laplacian are highly sensitive to microtopographic noise in the surface elevation map because the Laplacian uses the second derivative of the surface elevation. However, by the divergence theorem, the average Laplacian within any closed contour is equal to the integral of the normal gradient along the contour divided by the enclosed area. Therefore, we can examine the uniformity of the surface Laplacian by studying the slope of a regression between the integrated normal gradient and the enclosed area (Fig. 6). Indeed, we find a linear relationship between the integrated normal gradient along each contour and the area enclosed by the contour in our LiDAR-derived peat surface elevation map, indicating a uniform surface Laplacian in the region of uniform water table behavior (Fig. 6). In contrast, outside the region of uniform Laplacian, the water table behaves differently (“bog plain piezometer” in Figs. 5 and 6).

#### Uniform surface Laplacian determines stable tropical peatland morphology.

The uniform peat surface Laplacian provides a remarkably simple way to compute a stable morphology for a tropical peat dome. By “stable morphology,” we mean a morphology in which the peat surface and water table continue to fluctuate with the vagaries of climate, but there is no long-term average change in the peat surface or water table elevation (they are stationary; **7**) for the stable peat surface morphology **5**) with a trial Laplacian **3**) everywhere in the peatland (*SI Methods*). In this way, one finds a shape parameter (

#### Climate and drainage network determine tropical peatland carbon storage capacity.

By specifying the stable peatland topography, the uniform-Laplacian principle gives the peat carbon storage capacity inside any drainage boundary and in any given climate. The volume under the surface satisfying Poisson’s equation times the mean carbon density of the peat gives the carbon storage capacity of the peatland. For example, the peat dome at our primary site currently has a mean peat depth of 3.88 m (max 4.92 m) and stores about 1,535 metric tons (t) C

The uniformity of the stable peat surface Laplacian is an approximation that requires that (*i*) peat accumulation rate *ii*) flow of water is proportional to water table gradient (Boussinesq’s equation), and (*iii*) transmissivity is independent of location because flow through deep peat is negligible compared with near-surface flow. In reality, groundwater flow through deeper peat will result in a deviation of the stable peat dome surface from the uniform-Laplacian shape in very large peat domes. Specifically, groundwater flow through deep, low-permeability peat will tend to flatten the dome center, because of slow infiltration of water into the deep peat, and steepen the dome margin, because of exfiltration of water back into the high-permeability near-surface peat near the boundary. Deep groundwater flow should be manifested as a downward (dome center) or upward (dome margin) trend in the water table during nights without rain when the water table is low; no such trend is apparent in our piezometer data (Fig. 5*D*), suggesting that deep groundwater flow is small. A small deep groundwater flow term is further supported by radiocarbon dating of porewater dissolved organic carbon at our site (31), which suggests a maximum downward velocity of water of about 1 m/y or at most a 1.4-mm water table decline during a single 12-h night, 1/16th of the 22-mm water table decline from evapotranspiration during the day (Fig. 5). (Evapotranspirative flux is about 1/10th of the rate of decline of the water table from evapotranspiration because about 1/10th of the deep peat cross-section is available for water flow; see specific yield curve in Fig. 5*B*.)

A shape parameter related to our stable peatland Laplacian (Eq. **7**) appeared in Ingram’s model for temperate peatland morphology (10) assuming constant precipitation, uniform hydraulic conductivity, and simple river geometry (Ingram’s parameter is net precipitation divided by hydraulic conductivity, instead of average transmissivity). Our result is more general, because it handles varying rainfall and arbitrary landscapes, but is also mathematically simpler, because of our finding that transmissivity in tropical peatlands is approximately independent of peat depth.

### Dynamics of Tropical Peatland Topography and Carbon Fluxes.

#### Peat accumulation parameters regulate dome dynamics.

Our analysis shows how the rate of peat production **2**) have an indirect but strong effect on the stable peat surface Laplacian and hence peatland carbon storage capacity via the mean transmissivity **7**) because the mean water table depth must be equal to the ratio of the peat production rate to the decomposition rate constant (**3**). A higher decomposition rate constant implies a higher mean water table in stable peat domes, meaning a higher transmissivity, a smaller stable surface Laplacian, and less carbon storage. If both peat production

#### Fit parameters match literature data.

Peat accumulation parameters fitted to the topography of a peat dome at our Brunei field site agree with published data from other sites and also with our other field data (next section). We obtained peat accumulation parameters *A*; _{2} chamber measurements were also very similar to those from other sites, suggesting that the effect of water table on fluxes is similar at our site and in other tropical peatlands (Fig. 4*B*).

#### The uniform-Laplacian principle predicts a central bog plain and old peat near the surface at bog margins.

We find that a tropical peat dome reaches its stable shape first at its boundaries, because the stable dome surface is lowest there (Figs. 7 and 8 and Fig. S2). Meanwhile, the interior of the peat dome continues growing at an approximately uniform rate, forming a relatively flat (smaller-magnitude Laplacian) central bog plain. The vegetation of tropical bog plains may not be distinct (1), unlike the unforested bog plains of high-latitude peatlands (21); instead, we define the bog plain of a tropical peat dome as the central region that has not yet reached its stable Laplacian. Whereas the dome center continues to accumulate peat and sequester carbon, the margin has reached its stable shape and stopped growing, so peat near the surface is older there.

Older peat near dome margins has not been predicted before, so we collected 22 additional radiocarbon dates from basal and near-surface peat samples to test this prediction. These radiocarbon dates confirmed that near-surface peat was older near dome margins than at the same depths toward the interior of the same domes (Fig. 7*C*). We also compared radiocarbon dates in deeper peat to simulated ages at the same locations and depths, excluding basal samples from the mangrove peat before the establishment of the peat swamp forest (Fig. 7 and *SI Methods*) (1, 27). Radiocarbon dates and simulated ages at the same locations and depths matched well (Fig. 7*B*). We did not expect radiocarbon dates from cores to match simulated peat ages exactly because (*i*) the drainage network may have shifted during the 2,300 y of dome growth, (*ii*) tree root growth may inject young carbon into peat below the surface, and (*iii*) tree falls in peat swamp forests remove older peat to form tip-up pools that then fill with younger peat. In an earlier study, we estimated that replacement of older peat by younger peat in tip-up pools would bias radiocarbon dates of deep peat to about 500 y later than when material was first deposited in that stratum (figure 11 in ref. 27), consistent with the offset between measured radiocarbon dates and ages simulated by our model (Fig. 7*B*).

#### Carbon sequestration rate is proportional to bog plain area.

The centripetal pattern of dome development makes the rate of carbon sequestration roughly proportional to the area of the central bog plain (Fig. 8). Under a given climate, the rate of sequestration decreases as the dome approaches its stable shape and the central region of peat accumulation—the bog plain—shrinks in area. For example, our simulations imply that the current rate of CO_{2} sequestration at our site (0.80 t · ha^{−1} · y^{−1}, 100-y average) is less than one-quarter of its initial rate about 2,300 y ago (3.81 t · ha^{−1} · y^{−1}), and CO_{2} sequestration is more than five times faster at the dome interior (1.89 t · ha^{−1} · y^{−1}, 6.37 km from the river) than at its edge (0.36 t · ha^{−1} · y^{−1}, 1 km from the river; Fig. S3). The mechanism of tropical peat dome development that we describe therefore creates landscape-scale patterns in local carbon fluxes and radiocarbon date profiles. Local measurements of carbon fluxes or radiocarbon dates cannot be upscaled to regional fluxes without considering dome morphology because the flatter interior of each peat dome sequesters carbon whereas the margins do not (Fig. 8). Old peat near the peatland surface (2), although in some cases caused by local climate change or disturbance, also can be expected at the margin of any peat dome.

### Future Effects of Changes in Drainage Networks and Climate.

Our analysis provides a simple way of predicting long-term change in peat dome morphology and carbon storage in response to changes in drainage network, climate, or sea level because the stable peat surface Laplacian completely specifies the stable peat topography with given drainage boundary conditions. If the drainage network changes, we can solve Poisson’s equation in the new drainage boundary to compute the gain or loss of peat, and the net carbon emissions, as the peat surface approaches its new stable topography. If the climate changes, we can compute a new stable Laplacian value for the new climatic conditions and determine how much a currently stable peatland will grow or subside.

#### Subdivision of a peatland by drainage canals reduces carbon storage.

The average surface elevation of a stable peat dome is proportional to the area of the dome because of the uniform-Laplacian principle. If we scale the area of a peat dome by some factor

The dynamic response of a peat dome to changes in rainfall and sea level also depends on its area because of the centripetal pattern of dome development (Fig. 8). Because of their higher stable mean depth, larger-area domes reach their stable shape more slowly than smaller-area domes.

#### Relative effects of climate change on carbon storage capacity are independent of drainage network.

Although peatland drainage networks play a central role in determining absolute carbon storage and dynamics, we can calculate the proportional effect of climate change on long-term carbon storage of a tropical peatland independent of the drainage network. Poisson’s equation (Eq. **7**) must be solved in each drainage boundary to obtain the topography of the stable peat surface. However, we can then predict the effects of changes in climate independent of the drainage network because of the linearity of the Laplacian operator. By the definition of linearity for a mathematical operator, a peat surface Laplacian that is larger by some factor

#### Dynamic simulations converge to new stable morphologies after changes in conditions.

Our simulations of peat dome dynamics demonstrate the convergence of initially stable domes to new, stable, uniform-Laplacian morphologies after perturbations (Fig. 9). The simulations show the effect of increased total rainfall (Fig. 9 *A* and *E*), which is a recognized climate feedback for tropical peatlands (12), and also show that artificial drainage for agriculture (Fig. 9*D*) can dominate all natural feedbacks if not curtailed (4, 16). In addition, our simulations demonstrate a third feedback: The increase in rainfall variability from warming climates (33) can cause peat loss if not compensated by an increase in total rainfall (Fig. 9 *C* and *F*). For these simulations, we generated new rainfall time series as similar to current rainfall as possible but with larger annual and El Niño–Southern Oscillation (ENSO) fluctuations (Fig. S4 *A* and *B* and *SI Setting Annual and ENSO Amplitudes of Rainfall*). Either greater seasonality or a stronger ENSO decreased peatland carbon storage capacity, but an increase in seasonality had a larger maximum effect, partly because the magnitude of the ENSO fluctuation is smaller. In contrast, sea level rise could drive peat accumulation in the long term by elevating the tidal rivers draining most peat domes (Fig. 9 *B* and *E*). In general, losses can be much more rapid than accumulation (Fig. 9*E*), because subsidence of drained peatlands can be far faster than typical accumulation rates (4). For example, the estimated area-averaged current CO_{2} sequestration rate at our site is 0.80 t · ha^{−1} · y^{−1}, whereas Hooijer et al. (5) estimated CO_{2} emissions of at least 73 t · ha^{−1} · y^{−1} from tropical peatlands under plantation agriculture.

#### Intermittency of rainfall reduces tropical peatland carbon storage.

We find that fluctuations in net precipitation on timescales from hours to years can reduce long-term peat accumulation. We further explored the effects of variability in rainfall seen in our dynamic simulations (Fig. 9) by computing the effect of interstorm arrival time and annual and ENSO fluctuations on peatland carbon storage capacity (Fig. 10). The simulations demonstrate that long-term peat accumulation is controlled by variation in rainfall, not only by mean rainfall, because fluctuations in the water table cause exponential changes in groundwater flow. The high outward flow during peak water tables is not compensated by low flow rates after the water table declines. For example, a steady drizzle at the same average intensity as the intermittent rainfall actually observed at our site would sustain more than 10 times more long-term carbon storage (19.5 kt*D* and *E*). The intermittency of tropical convective storms significantly affects long-term carbon storage: Carbon storage capacity can decrease by one-third depending on whether convective storms arrive every 14 h on average, as at our site, or every 24 h, with the same mean rainfall (Fig. 10*A*).

Our simulations with smoothed rainfall intensity and evapotranspiration show that models must consider the effects of subdiurnal fluctuations in rainfall to correctly predict the long-term evolution and carbon storage of tropical peatlands. The exact details of the fluctuations in rainfall are not important, in the sense that many distinct rainfall time series can give the same stable surface Laplacian and the same carbon storage capacity. However, carbon storage capacity can be severely overestimated by simulations that entirely ignore the effects of fluctuations in rainfall. We explored the effects of neglecting fluctuations in rainfall by computing the stable surface Laplacian after averaging net precipitation on hourly and longer intervals. Treating rainfall intensity and evapotranspiration as constant each hour, instead of every 20 min, increased the simulated stable surface Laplacian by a few percent, but averaging over 1 d led to an overestimate by 20%, over 1 wk by 100%, over 1 mo by 400%, and over 1 y by more than 1,000% (Fig. S4 *D* and *E*).

## SI Setting Annual and ENSO Amplitudes of Rainfall

We explored the effects of fluctuations in rainfall on peat accumulation by changing the amplitude of annual and ENSO signals in rainfall time series. First, we evaluated the amplitude of annual and ENSO signals in our base rainfall intensity time series by simultaneously fitting the amplitudes and phases of annual and ENSO sinusoids plus a constant bias to daily average rainfall intensities. We chose a period for the ENSO signal of 7 proleptic Gregorian years (365.2425 d/y). We then generated new rainfall intensity time series for which the same fitting procedure would result in the same phases for annual and ENSO signals but with new, specified amplitudes. This is not as trivial as it sounds because the rainfall time series involves intervals of 1 mo and longer with no rain at all, so that simply adding a sinusoid to the rainfall intensity time series would result in (unphysical) negative rainfall.

We required that the new time series of *i*) had specified amplitude at the annual and ENSO frequencies and the same phase at those frequencies as the reference rainfall *ii*) was never negative; (*iii*) had the same pattern of storms as the reference rainfall, in the sense that if there was no rain in the reference time series, there was no rain in the new time series; (*iv*) had the same mean as the reference time series; and (*v*) was as close as possible in the least-squares sense to the reference rainfall while satisfying these constraints. Formally, this is the optimization problem

The least-squares cosine and sine coefficients

After fitting of the cosine and sine coefficients (Eq. **S38**), the mean of the residual **S44**) was then solved using a sparse QP solver (CVXOPT; cvxopt.org).

## Conclusions

The mathematical and numerical models presented here predict the long-term effects of changes in rainfall regimes and drainage networks on the morphology of tropical peat domes. Because tropical peat domes are mostly organic carbon, these predictions of peat dome morphogenesis also quantify peat dome carbon storage capacity and carbon fluxes. Our approach shows that tropical peatlands approach a limiting shape in which the Laplacian of the land surface is uniform. This stable peatland surface Laplacian can be computed from any rainfall time series and completely summarizes the effects of the rainfall pattern on the stable morphology and storage capacity of carbon within the peatland drainage boundary.

The uniform-Laplacian principle is supported by a range of observations: (*i*) The peat surface Laplacian is approximately uniform in a region near the dome edge (Fig. 6*C*); (*ii*) water table behavior is uniform where the surface Laplacian is uniform and is different in the dome interior (Fig. 5*A*); (*iii*) water table behavior is the same in areas with differing gradients within the uniform-Laplacian region (Fig. 5*A*); (*iv*) transmissivity increases exponentially at high water tables, so that local water balance is dominated by flow near the surface (Fig. 5*C*); and (*v*) peat accumulation parameters match literature data, even though those data were not used for calibration (Fig. 4*A*).

Our analysis underscores the importance of considering geomorphology when measuring and modeling carbon fluxes in tropical peatlands. On a growing peat dome, the perimeter of the dome reaches a steady elevation first while central areas continue to accumulate carbon (Fig. 8). This pattern of dome morphogenesis implies that the locations of ground-truth carbon flux measurements within tropical peat domes are important considerations for earth system models (34). For example, measurements of carbon flux in the center of a growing dome overestimate the average flux for the whole dome, because peat accumulation is fastest in the center (Fig. 8 and Fig. S3). The distribution of peat dome areas within a peatland complex is also important, because smaller domes reach their stable shapes faster after a change in conditions. Improved earth system models could use the uniform-Laplacian principle to efficiently account for the effects of changing rainfall, sea level, and drainage on tropical peat carbon storage, given a realistic distribution of peat dome sizes. The approach outlined here also provides a framework for including the effects of other long-term processes that remain understudied, such as shifts in river networks, changes in tree community composition, and saltwater intrusion from rising sea levels.

## SI Hydrologic Budget for Our Site

We can use water table and throughfall measurements to calculate an estimated water budget for our site (28). Evapotranspiration and divergence of groundwater flow can be distinguished in the decline of the water table between rain events because at low water tables, when divergence of groundwater flow is small, evapotranspiration creates a distinct diurnal pattern of steady water tables at night and declining water tables during the day (Fig. 5*D*). Therefore, after estimating specific yield from the ascending curve (Fig. 5*E*), we can estimate evapotranspiration from the daytime decline in the water table when the water table is low (Fig. 5*F*). Based on this estimate of evapotranspiration and our throughfall gauge data, there was 307 cm of throughfall to the peat at our site in 2012, of which 82 cm was lost as evapotranspiration from the peat. Overall precipitation falling on the peat forest and evapotranspiration from the forest are higher because some rain is intercepted by the canopy and evaporates before it reaches the peat surface. The increase in stored water in the peatland over the year, or recharge, was about 4 mm: The water table was 1.3 cm higher at the end of the data interval (January 20, 2013) than at the beginning (February 6, 2012), and the specific yield in that water table height range (5.2–6.5 cm) averages 0.3 (Fig. 5). Because the recharge term was only 4 mm, nearly all of the net precipitation (225 cm) was lost via divergence of groundwater flow.

## SI Master Recession and Recharge Curve Assembly

We assembled the master recession curve and master recharge curve by least-squares alignment of short water table time series representing intervals of either intense rain or no rain. We describe first the procedure for the master recession curve, which is slightly simpler. The head vs. time sequence is split into

Sets of time-vs.-head series can be assembled into a recession curve if one can navigate from the lowest to the highest head via regions of overlap between series. In most cases, there were some short sequences at the very highest and very lowest heads where a subset of series did not overlap with the rest. These small subsets of series were eliminated, leaving a single large connected component of time-vs.-head series. We also removed heads at which there was only one series because these contribute no information when finding time offsets.

We then solved for a time offset

We constructed master recharge curves using an approach similar to that used for master recession curves, based on the idea that when rain is intense, discharge is negligible by comparison

Rain storms did not arrive simultaneously at throughfall gauges and all piezometers, so it was necessary to match storms as recorded at throughfall gauges to the time of the storm’s arrival at each piezometer. We treated each contiguous interval of rapidly increasing head as a storm and then searched the throughfall time series for the corresponding record of that storm, considering that it might be slightly offset in time. Usually there was only one heavy rain that overlapped with the increasing head interval; in rare cases, there were two. In those cases we chose the storm with the higher total depth, assuming that it would be more likely to cause the rapid increase in head. If the time offset for onset of the storm differed greatly (by more than 1 h), the candidate storm was discarded.

Evapotranspiration estimated from the master recession curve represents loss from the connected porewater and surface water, not the total evapotranspiration from the forest. Total evapotranspiration also includes water intercepted by live foliage and water perched on the abundant leaf litter suspended above the peat surface and is probably considerably higher (28).

## SI Effects of Rainfall Aggregation Time

We explored the importance of short-term fluctuation in rainfall on long-term simulations of peat accumulation by experimenting with different time grids. In our simulations, we represented rainfall intensity and evapotranspiration as piecewise constant on 20-min intervals. Most models represent rainfall on a coarser time grid, making rainfall constant on timescales of months, years, or centuries, but we chose a much finer grid to represent rainfall intensity because our results showed that short-term fluctuations in rainfall can have a significant effect on the simulated long-term evolution of peat domes. We started with our 20-min time grid, on which rainfall is modeled as piecewise constant on each consecutive time interval:*t*′_{1} = *t*_{1}, *t*′_{n′} = *t*_{n}) but there are fewer total time steps (

## Acknowledgments

We thank Mahmud Yussof of Brunei Darussalam Heart of Borneo Center and the Brunei Darussalam Ministry of Industry and Primary Resources for their support of this project; Hajah Jamilah Jalil and Joffre Ali Ahmad of the Brunei Darussalam Forestry Department for facilitation of field work and release of staff; Amy Chua for logistical support; and Bernard Jun Long Ng, Rahayu Sukmaria binti Haji Sukri, Watu bin Awok, Azlan Pandai, Rosaidi Mureh, Muhammad Wafiuddin Zainal Ariffin, and Sylvain Ferrant for field assistance. We also thank Paul Glaser and two anonymous reviewers for their detailed comments on the manuscript. This research was supported by the National Research Foundation Singapore through the Singapore–MIT Alliance for Research and Technology’s Center for Environmental Sensing and Modeling interdisciplinary research program, by the US National Science Foundation under Grants 1114155 and 1114161 (to C.F.H.), and by a grant from the Environmental Solutions Initiative at Massachusetts Institute of Technology.

## Footnotes

- ↵
^{1}To whom correspondence should be addressed. Email: alex.cobb{at}smart.mit.edu. ↵

^{2}Present address: Gas Metrology Laboratory, National Metrology Center, Agency for Science, Technology and Research, 118221 Singapore.

Author contributions: A.R.C. and C.F.H. designed research; A.R.C. and J.E. established the field site; A.R.C., A.M.H., L.G., J.E., R.D., K.A.S., F.M.K., and N.S.H.S. performed research; A.R.C. contributed new analytic tools; A.R.C., A.M.H., and C.F.H. analyzed data; and A.R.C. and C.F.H. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Data deposition: The data reported in this paper have been deposited in the Dryad Digital Repository (dx.doi.org/10.5061/dryad.18q5n).

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1701090114/-/DCSupplemental.

Freely available online through the PNAS open access option.

## References

- ↵.
- Anderson JAR

- ↵.
- Dommain R,
- Couwenberg J,
- Joosten H

- ↵
- ↵.
- Couwenberg J,
- Dommain R,
- Joosten H

- ↵.
- Hooijer A, et al.

- ↵.
- Calle L, et al.

- ↵.
- Page S,
- Hooijer A

- ↵.
- Winston RB

- ↵
- ↵
- ↵.
- Glaser PH,
- Hansen BCS,
- Siegel DI,
- Reeve AS,
- Morin PJ

- ↵
- ↵.
- Gastaldo RA

- ↵.
- Clymo RS

- ↵
- ↵.
- Kurnianto S, et al.

- ↵.
- Morris PJ,
- Baird AJ,
- Belyea LR

- ↵.
- Baird AJ,
- Morris PJ,
- Belyea LR

- ↵.
- Jauhiainen J,
- Takahashi H,
- Heikkinen JEP,
- Martikainen PJ,
- Vasanders H

- ↵.
- Lampela M, et al.

- ↵.
- Glaser PH,
- Janssens JA

- ↵.
- Carlson KM,
- Goodman LK,
- May-Tobin CC

- ↵
- ↵.
- Ali M,
- Taylor D,
- Inubushi K

_{2}efflux from a tropical peatland in eastern Sumatra. Wetlands 26:612–618. - ↵
- ↵.
- Hirano T, et al.

- ↵.
- Dommain R, et al.

- ↵.
- Bonell M,
- Bruijnzeel LA

- Hooijer A

- ↵
- ↵.
- Baird AJ, et al.

- ↵.
- Gandois L, et al.

- ↵.
- Miettinen J,
- Shi C,
- Liew SC

- ↵.
- Huang P,
- Xie SP,
- Hu K,
- Huang G,
- Huang R

- ↵
- ↵.
- Moore ID,
- Grayson RB

- ↵
- ↵
- ↵.
- Ferziger JH,
- Perić M

- ↵.
- Patankar SV

- ↵.
- Moerman JW, et al.

- ↵.
- Partin JW,
- Cobb KM,
- Adkins JF,
- Clark B,
- Fernandez DP

- ↵.
- Moore JC,
- Grinsted A,
- Zwinger T,
- Jevrejeva S

- ↵.
- Roundtable on Sustainable Palm Oil

## Citation Manager Formats

## Sign up for Article Alerts

## Jump to section

- Article
- Abstract
- Methods
- SI Methods
- SI Expanded Description of Peat Dome Simulation
- Results and Discussion
- SI Setting Annual and ENSO Amplitudes of Rainfall
- Conclusions
- SI Hydrologic Budget for Our Site
- SI Master Recession and Recharge Curve Assembly
- SI Effects of Rainfall Aggregation Time
- Acknowledgments
- Footnotes
- References

- Figures & SI
- Info & Metrics