Atmospheric Rivers Feeding Terrestrial Floods in the Senyar 2025 Event: A Conceptual Review

Highlights

What are the main findings?

• The rare near-equatorial Cyclone Senyar depended on repeatable coincidences.
• Forest influence on windspeed may directly link to extreme rainfall.
• Orographic rain may involve windbreak effects before cooling of uplifted air masses.

What are the implications of the main findings?

• The rare coincidence behind Senyar may become more frequent in a changing climate.
• Flow-buffering effects of various land uses beyond natural forests need quantification.
• Lack of upstream water buffering causes flood damage to exposed people and homes.

Abstract

Before Cyclone Senyar made landfall on Sumatra in November 2025, it drew moisture from across the South China Sea, the Gulf of Thailand, and the Indian Ocean. The substantial damage its heavy rainfall caused on the NE and W coasts of Sumatra urges us to rethink the relationships between climate, forests, hydrology, land use, human presence, and vulnerability. Cyclones will recur, but flood damage does not have to be repeated. The simple narrative that “deforestation causes floods” is not adequate for guiding a “building back better” strategy. To account for the space-time pattern of Senyar effects with its multiple landfalls, we reviewed key concepts and framing of the links between ocean temperature, atmospheric moisture transport, rainfall extremes, saturation of existing buffers (“sponges”), river flow, and flooding. We hypothesize how atmospheric roughness slowing down rivers in the sky can induce congestion and precipitation. Surface infiltration and water retention in the soil profile matter. Mid- and downstream flow delays due to sponge effects depend on land cover beyond what a simple forest–non-forest terminology can represent. Rather than indiscriminate tree planting, adaptation efforts to avoid future disasters should balance reductions in human exposure (effective land use planning) and efforts to reduce hazards by restoring and managing vegetation and drainage systems.

1. Introduction

Protective (agro)forests can modify runoff genesis in mountain catchments in multiple ways. Most mechanisms follow a “downstream” highland-to-lowland logic. However, “downwind” impacts from lowlands to highlands may also play a role. The November 2025 floods in Sumatra were extraordinary not only for the severity of their impacts but for the meteorological conditions that triggered them. Warm surface water on the eastern side of the Indian Ocean coincided with warm La Niña conditions on the western side of the Pacific Ocean and in the Gulf of Thailand. This coincidence led to the interaction of two cyclones: Cyclone Senyar making landfall on the north-eastern coast of Sumatra and Cyclone Ditwa near Sri Lanka. Together, this drew in moisture from across the Indian Ocean. It produced exceptionally heavy rainfall over the western and northern coast of Sumatra. Impacts were felt in three Indonesian provinces: Aceh, North Sumatra and West Sumatra. Across these provinces, more than 1200 people died, approximately 80,000 people became displaced, and nearly 650 bridges were washed out, isolating mountain valley villages [1]. Rainfall maxima in climate station records were 311 mm/day in Aceh, 262 mm/day in Medan (N. Sumatra), 230 mm/day in Tapanuli, and 154 mm/day in W Sumatra. Damage was compounded by landslides [2]. Between the rescue and recovery phases of the disaster response, questions on “avoidable harm” and disaster attribution emerged, with multiple answers needing further analysis. “We didn’t deforest and still had landslides and floods to deal with”, a West Sumatra resident commented after the 2025 floods. Is all damage due to climate change or deforestation? Is tree planting as a default solution still relevant [3]? The interactions between forests and floods are complex and involve narratives across different scales and perspectives [4]. The special roles of forests, deforestation, and tree planting in the public perspective may exceed what scientific analysis corroborates.

Without claiming to provide answers to all questions and achieve a full attribution of the Senyar flood disaster, this study integrates diverse perspectives, priorities, and knowledge as found in the literature that is publicly accessible. It aims to help shape the questions and direction of future participatory research efforts. To achieve this, we address four research questions (RQ), as follows:

• RQ1. How does land cover (change) influence how cyclones and atmospheric flows become rivers on land?
• RQ2. How does land cover influence peak flow buffering and turbulent flows following peak rainfall events?
• RQ3. How have demographic shifts and land use change modified exposure to flood hazards in various parts of Sumatra affected by the Senyar floods?
• RQ4. What lessons can be learned for building back better?

The four interconnected research questions collectively describe the relationships among hazard, exposure, and vulnerability (Figure 1). RQ1 provides a basis for linking anthropogenic land cover change to the location and intensity of rainfall during a cyclone landfall. RQ2 takes the biophysical and ecological dimensions of hazard (e.g., rising water levels and turbulent flows) towards the socio-ecological dimensions of exposure to floods in specific locations in the watershed. Exposure can imply being buried by landslides in upper catchments or being affected by turbulent river flows and flooding in floodplains. RQ3 examines how socio-ecological dimensions of exposure relate to the social dimensions of vulnerability. Finally, RQ4 addresses the broader practical relevance of this analysis: reducing the likelihood of repeated damage and avoiding recovery efforts that simply restore a landscape that was already characterized by high levels of risk. Instead, it considers how recovery can contribute to a more effective build-back-better approach.

Figure 1. Conceptual hazard × exposure × vulnerability analysis of the impacts of the landfall of Cyclone Senyar on the N and W coasts of Sumatra in November 2025 and specific questions (RQs) for the current review of potential explanations.

Before exploring these questions for the specific case of Cyclone Senyar and Sumatra, basic concepts of atmospheric moisture transport, precipitation, and flood responses may need clarification with specific (internally consistent) hypotheses that deserve further testing. Some general background on the Indonesian climate and topography may help readers not familiar with the area.

2. Background on Atmospheric and Terrestrial Rivers and Forests

2.1. Rivers in the Sky

The terminology of atmospheric rivers first appeared in the modern scientific literature in the early 1990s. Language on “rivers in the sky” has had intuitive appeal in public discourse, including discussion of the rainfall patterns in the Amazon basin [5,6]. However, much of the scientific literature on moisture recycling over the Amazon and expected effects of deforestation appears to avoid the “river” terminology [7].

In the technical community, an “atmospheric river” is defined [8] as “A long, narrow, and transient corridor of strong horizontal water vapor transport that is typically associated with a low-level jet stream ahead of the cold front of an extratropical cyclone. The water vapor in atmospheric rivers is supplied by tropical and/or extratropical moisture sources. Atmospheric rivers frequently lead to heavy precipitation where they are forced upward—for example, by mountains or by ascent in the warm conveyor belt. Horizontal water vapor transport in the mid-latitudes occurs primarily in atmospheric rivers and is focused on the lower troposphere.” This definition is focused on laterally confined patterns of extra-tropical mid-latitude atmospheric moisture flux [9]. Other sources, however, set a threshold for integrated vapour transport (IVT) but do not specify how long and narrow a stream must be to qualify as a river. They also treat latitude as an empirical issue of where specified phenomena occur, rather than as part of the definition.

Applicability of the “atmospheric rivers” concept in the tropics is still debated. Atmospheric river detection tools either use an absolute threshold (e.g., IVT > 250 kg m⁻¹ s⁻¹) or a relative threshold based on the climatology of the proxy (e.g., IVT > 95th percentile of climatic records for the region) [10]; a catalogue of Atmospheric River Trajectories contains an equatorial area on the eastern side of the Indian Ocean (to the W of Sumatra) and a number of mid-latitude areas.

A gauging and modelling study of atmospheric rivers [11] found that existing models of short-range water vapor flux (the atmospheric branch of the global hydrological cycle) had a root-mean-square error of 22% of the mean observed flux. Authors attributed overall uncertainties primarily to those in the low-level winds, where vegetation impacts may be strongest. Greater predictive skill of models is desirable. Beyond stronger statistical databases, it may require more process-level understanding than is currently available, as events are rare. The effects of land cover change (including ‘deforestation’) are particularly policy relevant, as they may point to actionable aspects of damage prevention. Deforestation can substantially reduce surface roughness and increase windspeeds [12,13], extending the distance of moisture transport but potentially also reducing local precipitation and causing part of incoming atmospheric moisture to leave a focal basin without recycling.

Rivers in the sky primarily affect coastal regions. However, flooding on the ground in the lower part of the Rhine catchment area in central Europe is also linked to high IVT events. On the days preceding flood peaks, convergence of prevailing atmospheric flows was noted at lower levels of the atmosphere and divergence in the upper levels, indicating strong vertical motions and heavy rainfall [14]. Vertically integrated water vapor transport (IVT) exceeded 600 kg m⁻¹ s⁻¹ for the largest floods. Integrated vapor transport (IVT)—with kg m⁻¹ s⁻¹ as units [15]—is the integral over a vertical column of atmospheric moisture or total precipitable water (TPW) (kg m⁻² or mm) and velocity (m s⁻¹). An atmospheric column can contain up to 70 mm (or kg m⁻²) of ‘total precipitable water’ (TPW) at any point in time [16,17]. This means that a windspeed of 3.6 m s⁻¹ (12.9 km h⁻¹) can be sufficient to match the 250 kg m⁻¹ s⁻¹ IVT threshold for atmospheric rivers.

The mean residence time for atmospheric moisture of 8–10 days with a median of 4–5 days [18,19,20] indicates a long-tailed statistical distribution. Residence times over the ocean are about 2 days less than those over land [21], reflecting TPW values that are closer to their temperature-dependent saturation—TPW_max(temp)—levels. The distance travelled in the atmosphere can be derived as velocity divided by residence time: 9 days at 1 km h⁻¹ implies a distance of 216 km, at 10 km h⁻¹ 2160 km, and at 100 km h⁻¹ during a storm, this will be 21,600 km, or of the order of half a continent. More sophisticated models with more than one layer in the atmosphere with different windspeeds and temperatures can refine predictions on the ‘short cycle’ of water over land [22].

The ‘rivers in the sky’ follow physical rules that differ from rivers over land as they are not physically confined and are not limited in the velocity that can be attained. The maximum storage capacity of atmospheric moisture, however, is constrained by temperature, and where it is exceeded, precipitation is the primary way out. Mainstream explanations of precipitation, however, focus on other causes of ‘convergence’.

2.2. Convergence and Congestion as Central Concepts

In the Inter-Tropical Convergence Zone (ITCZ) [23] surface winds from the south and winds from the north meet and converge, causing upward vertical movement of atmospheric moisture and precipitation. The upward movement implies cooling and precipitation. Higher-level atmospheric flows move air masses away from the ITCZ in ‘Hadley cells’. Convergence of streamlines where mountain ranges influence flow pathways can also cause conditions where the inflow of moisture into any volume element (or ‘voxel’) in the atmosphere exceeds the outflow. In the climate literature, any situation where the gross import exceeds the export of water vapour in a location of interest is described as convergence and linked to precipitation [24]. However, just as both the convergence of lanes in a highway causes a bottleneck, and rough road sections or traffic lights that slow down traffic can cause congestion, atmospheric congestion (where inflow exceeds outflow) may be due to changes in velocity as such, rather than to intersecting streamlines (Figure 2).

Figure 2. Convergence and congestion concepts: (A). Convergence of streamlines on a two-dimensional map (B). The Inter-Tropical Convergence Zone (ITCZ) that moves with the seasons and involves vertical and horizontal air movement in ‘Hadley cells’ (C). Reduction in one-dimensional windspeed (D). Definition of congestion (potentially induced by convergence) as a condition where inflow of atmospheric moisture into a voxel exceeds outflow and precipitation is a way out where maximum (temperature-dependent) storage capacity has been reached.

Two hypotheses emerge here: ‘windspeed reduction causes congestion’ and ‘congestion causes rainfall’. Combined, these hypotheses imply that lowland coastal forests, by slowing winds and inducing rainfall, can protect hinterland areas from the full impact of ocean–land atmospheric moisture transfers. Published evidence supporting or rejecting these hypotheses is scarce. Quantification may (at least) test their internal consistency.

The IVT flux of atmospheric moisture at any point of its trajectory equals the (suitably height averaged) velocity (V_atm, ranging from 0 to >100 km h⁻¹) times the total precipitable water content (TPW, ranging from 0 to >80 kg m⁻²). Conservation of the mass balance of atmospheric moisture implies that a gradual reduction in V_atm will lead to an equivalent increase in TPW, until a temperature-dependent TPW_max(temp) is reached and precipitation P is triggered. Concurrent evapotranspiration E can replenish the atmospheric moisture flux.

In algebraic form, the product rule of calculus (d(Y × Z)/dx = Y × (dZ/dx) + Z × (dY/dx)) implies that a spatial gradient Δ() in IVT as the product of V_atm and TPW can be split into two components: the spatial gradient in V_atm multiplied by the average TPW and the spatial gradient in TPW multiplied by the average V_atm.

P − E = Δ(V_atm × TPW) = V_atm × Δ(TPW) + TPW * Δ(V_atm)

(1)

and two complementary contributions to the P−E estimate:

RelShare_V_atm_change = V_atm × Δ (TPW) /(V_atm × Δ(TPW) + TPW × Δ(V_atm))

(2)

and

RelShare_TPW_change = TPW × Δ(V_atm) /(V_atm × Δ(TPW) + TPW × Δ(V_atm))

(3)

More sophisticated versions of theory will need to include turbulent flows and multi-layer feedback loops [25] and consider error propagation where empirical data are used.

Popular accounts of ‘orographic rainfall’ emphasize cooling effects of vertical shifts in airflow with consequences for TPW_max(temp) and appear to ignore the potentially relevant ‘windbreak’ effects of mountains, reducing V_atm before the mountains are reached. The reduction in windspeed, and, hence, rainfall, may occur before the actual mountain is reached, like other ‘windbreak’ effects. Equations (2) and (3) suggest that the relative change in V_atm due to ‘congestion’ can be manifold larger than the change in TPW_max(temp) in the conventional explanation.

This simple description is focused on atmospheric moisture treated as the non-reactive component of the atmosphere. Concomitant changes in ‘air pressure’ based mostly on non-water components of the atmosphere interact with the dynamics of windspeed in complex feedback loops. Pressure differences can determine the direction of atmospheric moisture flow, with further ‘congestive’ conditions in the curves of flow paths. Condensation and evaporation as phase shifts cause further complications for air pressure.

2.3. Contrasting Rivers in the Sky and Rivers on Land

Where precipitation is in the form of rainfall (rather than snow that waits for conditions facilitating snowmelt), it can be absorbed by aboveground vegetation, stored in the (non-saturated) soil or reach streams and rivers (Figure 3). The water flux in rivers equals V_riv × W × H, where width (W) and water level (H) are determined by local geomorphology, with potential human modifications. Where W is constrained in a riverbed (or even more so in a canal), an increase in H and possible bank overflow are the main degrees of freedom to increase flux. The velocity V_riv increases modestly when H and flow volume increase as it is determined by the constant elevational gradient of the river, modified by surface roughness (as characterized by the Manning factor) that decreases with flow volume.

Figure 3. A conceptual understanding of how the atmospheric part of the ocean–land–water cycle transports precipitable water, generates rainfall, and differs from rivers on land that are confined to channels with defined width and height that flow back to the ocean if not atmospherically recycled.

Other than a ‘congestion’ effect in atmospheric flux that leads to P, a reduction in V_riv when the river enters peneplains tends to be compensated by an increase in W, in what are often described as floodplains. In these floodplains, V_riv is reduced and river-born sediments can be deposited (first stone fractions and sand, followed by silt and clay). As the sediment-carrying capacity of a river involves V_riv to the power 4 (by approximation), relatively small changes in V_riv can induce sedimentation. Sedimentation in riverbeds and floodplains remains vulnerable to a next flood event with higher V_riv levels, unless vegetation with superficial root development stabilizes it. Considering these essential differences in the determinants, constraints and degrees of freedom of rivers in the sky (generating P) and rivers on land (transporting non-buffered P), the ‘river in the sky’ language may be appreciated for its poetic power but may mask essential differences.

A recent review of the published evidence for a forest-based reduction in destructive tropical cyclones, hurricanes and typhoons [26] distinguished between potential influences during A. cyclone formation over (warm) ocean water, B. rainfall patterns during landfall and C. subsequent flow dynamics of terrestrial rivers; it used the recent Senyar cyclone as the trigger.

2.4. Forests and Rainfall

In the global assessment of forest–water relations [27], a shift between two generic paradigms (‘all forests are good for all hydrological functions’ and ‘forests use more water than other vegetation’) was a prelude to a synthesis that emphasizes the importance of location, especially for downwind rainfall effects [28]. The new synthesis suggests a quantitative tree cover continuum rather than a dichotomous forest/non-forest classification. The broad range of land use systems that include trees [29] at a wide range of landscape positions influences the water balance, regardless of any specific forest definition used.

Atmospheric moisture represents only 0.001% of the amount of water on planet earth and 0.04% of global freshwater but is a powerful driver of global weather and climate. As a global average, about 60% of rainfall over land originates from terrestrial evapotranspiration [30], varying from nearly 0 at the coast to nearly 100% in the centre of continents [31]. The rediscovery of the full hydrological cycle (ocean–land and land–land) brings climatological, hydrological and ecological science of droughts, floods and intermediate ‘normal’ water supply conditions together with a range of social and economic sciences. Droughts and floods are not ‘just’ extremes of a statistical distribution of water availability but triggers of ecological and human adaptive responses and cascading disasters where past adaptation has been insufficient [32,33]. The increasing popularity of the ‘rivers in the sky’ terminology for flows of atmospheric moisture calls for a comparison with the behaviour of rivers on the earth’s surface.

Quantitative evidence exists for forest and tree roles in increased surface roughness during landfall of 19 cyclones that hit Australian coasts during the period 1984–2010; surface roughness and central pressure (proportional to windspeed at landfall) yielded a combined coefficient of determination of 76% during calibration and 59% during validation [34]. Like the buffering effect of coastal tree vegetation on incoming tsunami waves [35], moisture-laden winds lose water after landfall when they are slowed by surface roughness. Direct translation of the Australian study to Sumatra may need to incorporate differences in underlying topography and differences in tree cover types, especially where roughness depends on variations in tree height, rather than tree height as such.

2.5. A Simple Atmospheric Moisture Flow Model

A one-dimensional model of atmospheric moisture represents transport perpendicular to a coastline. An incoming amount of atmospheric moisture (TPW × V(0) × T) reaches the land (e.g., TPW 65 mm, V(0) = 40 km h⁻¹, T = 6 h). If this would fall evenly over a coastal zone of 150 km, rainfall depth would be 104 mm for any station. However, terrain roughness plus vegetation effects will influence a ‘windspeed persistence’ across adjacent cells. For our conceptual model, we considered cells of 30 km length and—as illustrative assumptions—a windspeed persistence of 0.33 for terrain of high roughness and a persistence of 0.67 for terrain of lower roughness. We assume windspeed due to other causes to have a minimum value, 0.5 km h⁻¹ in our numerical example. Resulting windspeeds would be 13.2, 4.4, 1.4, 0.5 and 0.5 km h⁻¹ at 30, 60, 90, 120 and 150 km from the coastline for the high-roughness case, and 26.8, 18.0, 12.0, 8.1 and 5.4 for the low-roughness case, respectively. Resulting station-level rainfall depths would be 348.4, 119.3, 37.9, 10.9, and 1.0 mm for the high-roughness case, and 171.6, 132.9, 95.2, 63.9, and 29.4 mm for the lower-roughness case, respectively. A reduction in roughness due to deforestation would decrease rainfall in the first cell but increase it further from the coast (Figure 4), leading to inconclusive deforestation impacts on station-level rainfall, despite influencing relevant processes.

Figure 4. Results for a simple conceptual model where surface roughness induces a reduction of V_atm for a front travelling over land from the coast, with consequences for windspeed and total precipitable water (TPW (left panel) and the spatial distribution of precipitation (P)− evapotranspiration (E) (right panel).

Within say 100 km from the coast, nearly all TPW is expected to become rainfall. In the first zone along the coast, the restored presence of forest would increase rainfall, but further from the coast, rainfall will be higher if coastal zones have lower atmospheric roughness (Figure 4). If this is a dominant pattern, a simple regression of rainfall on local forest cover will probably not show statistically significant relationships, as has been the general conclusions for studies based on station data. It suggests that where lowland areas near the coast became flooded, local rainfall was a major contributor, rather than river flow from upper and middle parts of the watershed. Specifically for the N Sumatra area, the relationship between rainfall and deforestation was extensively studied in the 1920s when large-scale plantations were rapidly expanding, especially on the east coast—but data for that episode analysed in reference [36] could not reject a ‘no impacts’ null hypothesis, as space-time variability was substantial. Analysis of data obtained over a period of more than 40 yr, for each month, at each meteorological station in Thailand, a country with substantial deforestation during the years the data refer to, revealed significant decreases in precipitation over Thailand only in the time series of monthly precipitation in September [37]: a decrease of approximately 100 mm month⁻¹, a 30% relative change. A well-cited review of forest–water interactions [38] concluded that effects on rainfall of forest disturbance and conversion are smaller than the average decrease of 8% predicted for a complete conversion to grassland in southeast Asia, while the higher evapotranspiration and greater aerodynamic roughness of forests compared to pasture and agricultural crops will lead to increased atmospheric humidity and moisture convergence and, thus, to higher probabilities of cloud formation and rainfall generation, empirical data report both confirmation and rejections of hypothesized generic effects. A conceptual model that can account for both increases and decreases in rainfall depending on changes in the ‘cloud stripping’ and wind persistence impacts of atmospheric roughness in upwind forests will need further empirical tests but is conceptually relevant.

2.6. Flow Buffering

Given a spatial pattern in rainfall, river flow will concentrate rainfall excess in a specific channel that is often a primary attractor of human activity. In a catchment where all soil surfaces are sealed and no water infiltrates the soil, as happened during a certain phase of ‘modernization’ of cities, the drains, streams and rivers must deal with all rainfall instantaneously. Where there are vegetation and a living soil, part of the rainfall is (Figure 5):

Figure 5. Schematic partitioning of a peak rainfall event over temporary on-site and in-landscape storage versus flow pathways that reach the streams and rivers within 24 h with further dynamics dependent on the river network and bank overflow opportunities.

• Intercepted by the leaves of plants;
• Captured in a surface litter layer that protects the soil;
• Rather than flowing off over the soil surface, infiltrates the soil, especially where active soil life, such as earthworms, maintains soil porosity [17,39];
• At field scale, some of the overland flow can be trapped due to surface roughness and local ponding;
• The water infiltrated will first replenish soil moisture absorbed by plant roots after the previous rainfall event;
• The surplus water can normally (with exceptions causing saturation overflow) find its way through the soil profile to a downhill riparian zone on sloping land and/or vertically replenishing groundwater that can gradually seep into rivers.

The processes can be quantified based on process-based research in various land cover types. The quantities of water involved in each of these processes depend on the timing and intensity of a rainfall event, of features of the terrain, the inherent characteristics of soil, the vegetation and human impacts on the soil–plant–atmosphere system. The amount of rainfall that does not show up in the river within one day of a rainfall event is said to be ‘buffered’. This will be close to zero for the city and may be above 90% for undisturbed natural forests.

A flow persistence (F_p) or buffering indicator links two aspects of water retention: reductions in peak flow after precipitation events and a gradual release of stored water maintaining base flows. Temporal autocorrelation in daily river flow data ((Q_t, Q_t+1) pairs) can be explored to derive the F_p parameter according to [40]:

Q_t = F_p Q_t−1 + (1 − F_p) (P_tx − E_tx)

(4)

where P_tx is the (spatially weighted) precipitation on day t (or preceding precipitation released as snowmelt on day t) in mm day⁻¹; E_tx, also in mm day⁻¹, is the preceding evapotranspiration influencing antecedent soil moisture that allowed for infiltration during this rainfall event (i.e., evapotranspiration since the previous soil replenishing rainfall that induced empty pore space in the soil for infiltration and retention). Between the extremes of natural forest and sealed city, a wide range of ‘buffer coefficients’ (between 0 and 1) can be found.

Calculations on river peak flow can also be made using the ‘curve number’ approach that originated in the USA [41]. Where there is a well-established empirical base, the curve numbers can be estimated for all relevant land cover types in combination with soils and terrain. Where such data are scarce, however, a more process-based approach to link ‘structure’ of vegetation and soil to function may be more flexible as it can deal with intermediate land cover types, such as found in ‘agroforests’, a category relevant in Indonesia but not in the USA.

A process-level understanding suggests that the value of F_p will depend on P_tx and that under high rainfall, the buffer capacity is saturated once the landscape is water-filled, but in the data series analysed so far, such saturation was not evident [42]. In land use scenarios that—based on research in Sumatra and other locations—can be expected for Sumatra, the difference between F_p values for a natural forest-dominated landscape and one affected by logging (reduced interception, partial soil compaction, availability of tracks as surface drainage channels) may be around 30% of the (rainfall-dependent) background value. Buffering in agroforest landscapes (with patches of rice paddies in the valleys) may be like that in natural forest recovering from logging, where a further 10% reduction can be expected for forestry, rubber or oil palm plantation landscapes. Intensified open-field agriculture and settlements will have an F_p below 40% of that for natural forest. Area-specific parametrization of the GenRiver model [42,43] can refine these estimates. The buffering indicator for paddy rice terraces is like that measured for multistrata agroforests [44] but based mostly on the surface storage term of the relevant equations.

2.7. Forest Concepts

Discussions on the way environmental problems interact with changes in the quality and quantity of ‘forest’ in Indonesia (and elsewhere) are hindered by a lack of shared understanding of what non-forest is (Figure 6). As a social-ecological boundary concept, the term forest in Indonesia has a strong institutional meaning (the ‘designated forest’ or ‘kawasan hutan’ typically claimed by the State) and a biophysical one centred on tree cover and quantifiable ecosystem services and hydrological relationships. Functional distinctions in tree-covered lands with various intensities and degrees of agricultural use dominate interactions with the water balance (Figure 6). In our analysis of ‘hazard’, we focus on the relevant functional traits, while for vulnerability and ‘building back better’ discussions, the institutional distinctions may matter most.

Figure 6. Biophysical and institutional interpretations of the term ‘forest’ that tend to interact in popular debates over ‘deforestation’ as a specific form of land use and land cover change [4].

The current debate interfaces generic theory on how forests are different from all other types of land cover, with the possible idiosyncrasies of a specific cyclone in each location. Forest, as a high-level category without specifying the structure, function and location of what was observed, is challenged by the low degree of consensus on how much forest is left on the globe and where it is. This calls for a more detailed land cover classification in testable hypotheses. A recent comparison of ten global maps showed consensus on a forest status for only 26% of the pixels identified as forest in at least one dataset [45]. Similarly, the challenges in making policies operational that are to secure a deforestation-free status of traded commodities [46,47] cast doubt on the continued use of a forest–non-forest dichotomous classification in environmental policies. A recent study in East Java showed a substantial difference in infiltration rates on steep volcanic slopes between remnant natural forest and Pinus merkusii plantations [48]. Differences within the ‘forest’ category are important for functional attributions.

2.8. Flood Frequency

In discussing floods with various stakeholders, it may help to clarify the terminology for the various time and spatial scales involved in ‘floods’ (Table 1). In policy discussions, the distinctions between the last three categories (once a decade, century or millennium) matter for the (implicit) tolerance of risk in view of financial and opportunity costs involved in reducing expected damage.

Table 1. Differentiation of flood phenomena from a user perspective with potential responses and required technical expertise.

Frequency	Context	Adaptive Land Use	Responses	Technical Expertise
Twice a day	Tides in coastal zone	✓ Fishing, Mangrove management	❖ Assisted man-grove regeneration	❑ Coastal-zone manager
Once a year	Seasonal floodplains in peneplains and delta’s	✓ Grazing, Wildlife ✓ Seasonal crops on floodplains	❖ Wetland management ❖ Zoning human use	❑ Forest-hydrologist ❑ Hydrological engineer ❑ Agronomist/soil scientist
Once per decade (0.1 year−1)	Exceeding naturally formed ‘bank-full’ flow beyond seasonal floodplains	✓ Flood-tolerant tree crops, ✓ Seasonal crops, ✓ Temporary housing (ready to abandon)	❖ Dykes & embankments ❖ Space for river ❖ Clean urban flood canals	❑ Eco-hydrologist ❑ Engineer (drains, dykes, embankments) ❑ LU planner ❑ Insurance banker
Once per century (0.01 year−1)	Maps of the riverine or coastal 100-year floodplain inform building permits, and flood insurance	✓ Agriculture, plantations ✓ Low-cost (risk tolerant) housing	❖ Space for river ❖ Drainage/irrigation ❖ Select emergency inundation zones ❖ Climate change	❑ Urban planner ❑ Engineer (drains, dykes, bridges, roads)) ❑ Insurance banker ❑ Climate change analyst
Once in millennium (0.001 year−1)	Acceptable for all but most sensitive urban uses	✓ Non-risk tolerant housing and industry	❖ Urban planning ❖ Climate change adaptation plans	❑ Urban planner ❑ Engineer (drains, pumps, buildings) ❑ Insurance banker ❑ Climate change analyst

3. Background on Indonesian Topography and Climate

Indonesia is the world’s fourth most populous country. It is a tropical archipelago of some of the world’s largest, many smaller and thousands of uninhabited, islands, located between Indian and the Pacific Ocean. As part of the ‘ring of fire’, volcanically rejuvenated soils are common and have historically allowed for high human population densities to emerge, especially in Java and, to a lesser extent, on Sumatra. The eastern parts of Indonesia are in the rain shadow of the Australian continent, especially in the June to September period, when monsoons come from the south and east, while they come from the north-west in the December to March period. Prevailing wind patterns interact with local topography to generate significant variations in rainfall throughout the archipelago.

Most of the rivers in Indonesia are relatively short, connecting mountain ranges to the nearest ocean. Relatively long (and large) rivers are found to the east of the Bukit Barisan mountain range of Sumatra, on the island of Borneo and Papua and on the N coast of Java. A combination of high tectonic and volcanic activity and short rivers can explain the exceptionally high sediment load of rivers, with active mangrove formation where sea currents allow. The floodplains formed by these geological processes offer attractive opportunities for agricultural expansion, once seasonal flooding patterns are taken onto account or controlled by specific drainage interventions. The middle and lower reaches of the rivers tend to have ‘energy-limited’ sediment transport, implying that a substantial increase in flow rate and water level can mobilize large volumes already in the river channel, rather than requiring concurrent erosion [4].

3.1. Climate

Rather than by frontal rain patterns derived from the ocean, precipitation in the Indonesian archipelago (sometimes called the Maritime Continent) is primarily driven by a very strong diurnal cycle of local convection that adds ocean-derived to locally recycled atmospheric moisture. However, seasonal and interannual variations in sea surface temperatures matter. Three distinct climatic regions were described for Indonesia [49], with bimodal rainfall patterns in the northern part of Sumatra and Kalimantan and unimodal patterns with a single rainy and dry season in the rest of the country. Analysis of Indonesian rainfall data for 1985–2010 in relation to Indian Ocean Dipole (IOD) and the El Niño/La Niña cycle [50] showed that rainfall in north-western Sumatra was positively correlated with a positive IOD value, while in southern Sumatra and Java, it correlated with a negative IOD value; in eastern Indonesia, rainfall was positively correlated with La Niña, while in central Indonesia, seasonal variations due to monsoons were predominant. The Indian Ocean contribution to rainfall in Sumatra can be linked to convectively coupled Kelvin waves [51] traveling from west to east along the equator at around 40 km h⁻¹ (∼12 m s⁻¹). Most analysed floods in Sumatra could be linked to landfalls of such equatorial Kelvin waves [52]. Cyclones (also known as hurricanes or typhoons) depend on Coriolis forces and are very rare within a 5° N to 5° S belt around the world.

3.2. Land Use Patterns

An analysis of historical demographic data for Sumatra starting in the 19th century ([53]; Figure 7) suggested an interesting reversal from common expectations. Elsewhere in Asia, migrations of lowlanders into sparsely settled highlands dominated, but in Sumatra, it was the reverse: people moved from their relatively crowded and impoverished high-land valleys into the coastal cities and lowland planes when these became sufficiently safe. The lowland valley and deltas were not in fact hospitable. Floods were a constant problem. Only large-scale irrigation and drainage works could control the large volumes of water in the lowlands. The percentage of the total population of West Sumatra living in coastal lowlands increased from 9 around 1830 to 20 by the 1850s, 36 in 1920/30 and 49 in 1990. For North Sumatra, equivalent estimates are 21% in the 1850s, 67% in 1920/30 and 77% in the 1990s.

Figure 7. Historical demographic concentrations in the uplands of Sumatra based on reference [53].

For Aceh, a long history of conflict before and after Indonesian independence coloured the inland–coastal zone relationships. Inland valleys were buffered from outside forces arriving by sea, while state-based security could only be guaranteed in coastal zones. The 2004 Tsunami hit the coastal zones and shifted the balance of power in the province, allowing a political settlement of past conflicts. Lowland populations that recovered from the Tsunami were badly affected by the November 2025 floods, while inland populations were once again cut off from the coast when roads and bridges washed out. The political backgrounds of different types of flood vulnerability need to be appreciated before effective recovery and avoidance measures can be designed.

4. (How) Does Land Cover Influence the Way Cyclone Rivers in the Sky Become Rivers on Land (RQ1)?

The damage Senyar caused in the northern parts of Sumatra was exceptional in a near-equatorial area where cyclones do not normally develop. The affected areas lost a considerable part of their natural forest cover in the past decades. Causal effects of change in forest cover remain contested. Senyar caused apparently unprecedented damage to intact forest on the Bukit Barisan mountain range, especially in the Batang Toru landscape (North Sumatra), home to the critically endangered Tapanuli orangutan (Pongo tapanuliensis) [54].

In the specific case of the Senyar cyclone, a rare coincidence of a La Nina phase in the Pacific Ocean with warm ocean waters to the east of the Indonesian archipelago and a negative Indian Ocean Dipole with warm ocean waters to the west of the Indonesian archipelago led to warm waters in the Gulf of Thailand and a flow of moist air that crossed the Malaysian peninsula at its narrowest point near Songkhla in S Thailand (Figure 8). A relatively low surface roughness due to forest conversion at the Malaysian peninsula may (further quantification and analysis is warranted) have contributed to the apparent ease of moist winds to cross over from the Gulf of Thailand and to the addition of atmospheric moisture feeding the start of the Senyar Cyclone around November 24. As a parallel cyclone nucleus developed around Sri Lanka, the Senyar circulation could feed off Indian ocean flows and hit Sumatra’s W coast, with W Sumatra and Tapanuli hit hardest on the SW side of the Bukit Barisan range with rivers flowing towards the coast overflowing their banks. Subsequent Senyar landfall was around the border between Aceh and N Sumatra province and the lowlands here, with much of previous mangrove converted to rice paddies flooded and receiving silt and clay deposits that damage crops (in the short run) and flood housing areas (built on floodplains). Strong winds and rainfall hit the NE side of the Bukit Barisan range, with rivers flowing towards the coast overflowing their banks.

Figure 8. Atmospheric conditions in November 2025 when the Senyar and Ditwa cyclones interacted as represented on the https://earth.nullschool.net reference website (accessed on 15 August 2026) for the specified location and date. The data in the lower-left insert refer to the green circle sample location.

For the ocean-to-land moisture transfer associated with cyclone landfalls, the windspeed and TPW (total precipitable water, expressed in mm and generally below 70 mm) determine the potential rainfall, where windspeed is reduced and TPW would otherwise exceed the maximum values rainfall occurs. Point-level rainfall on land can be multiple-times the maximum TPW_max(temp), and values recorded in Sumatra up to 350 mm in 24 h are possible. Windspeeds at landfall may still be around 30 km h⁻¹, and where they are reduced to around 5 km h⁻¹, six-times TPW is possible as rainfall. Following this simple logic (mass balance for atmospheric moisture) with values for windspeeds and TPW encountered in the Senyar context, we can explore the effects of vegetation roughness on the spatial pattern of rainfall (e.g., by a high and low value).

A more detailed account of the day-to-day dynamic in November 2025 (Figure 9) shows that Senyar, after forming in the Strait of Malacca, appeared to be heading for the Malaysian peninsula on 23 November before connecting with cross-Sumatra atmospheric flows and hitting the coast of East Aceh and North Sumatra on 26 November. The strong winds reached the W coast of Sumatra from 24 to 27 November, shifting southwards, and were probably stronger than the normal equatorial Kelvin waves. For a transect perpendicular to the coast, the dynamic of TPW and V_Atm (Figure 10) revealed interesting patterns.

Figure 9. Day-to-day dynamics of atmospheric conditions over and around Sumatra in November 2025 as represented at the https://earth.nullschool.net website; the green circle indicates sample point 5 in Figure 10.

Figure 10. (A). Sample locations (1…7) in a transect perpendicular to the coast in West Sumatra, approximately aligned with prevailing winds for 8 days (20…28) in November 2025. (B). Surface windspeed and precipitable water (TPW) data derived from the reference in Figure 8 and hourly estimate of P−E derived from the spatial gradient in atmospheric moisture transport (TPW × V_atm).

The observed IVT flux (Table 2) reached values above 500 kg m⁻¹ s⁻¹ to the W of the Sumatra coastline but had dropped to values below 270 kg m⁻¹ s⁻¹ at the coastline itself, and values below 100 kg m⁻¹ s⁻¹ within 100 km inland from the coastline. In the last 50–100 km before reaching the coastline, the windspeed was already reduced by half (Table 2, Figure 10) with further reductions in the first 50 km of overland travel—essentially before the Bukit Barisan mountain range was reached. Estimates of P−E based on the spatial gradient in the (TPW × V_atm) product suggest heavy rainfall in the coastal zone, on both sides of the ocean–land transition. This pattern is consistent with satellite-derived long-term rainfall data provided by Baranowski et al. [52] (Figure 11).

Table 2. Data for windspeed and total precipitable water (TW) for the 7 locations in Figure 9 for 9 days in November 2025 derived from the https://earth.nullschool.net/ (accessed on 15 August 2026).

Distance to Coast, km	Date (Nov. 2025)	20	21	22	23	24	25	26	27	28	Avg.
	Windspeed, km h−1
−450	1	12	18	16	18	23	28	30	27	18	21.1
−300	2	14	18	11	24	19	25	31	32	20	21.6
−150	3	13	11	13	21	16	14	22	27	15	16.9
−75	4	11	13	16	24	23	15	25	31	16	19.3
0	5	5	7	12	14	15	5	13	13	7	10.1
50	6	6	5	9	9	7	3	4	4	2	5.4
100	7	3	5	8	6	3	6	7	6	4	5.3
	TPW, kg m−2
−450	1	59.7	53.7	56.9	53.5	58.3	48	55.2	56.3	61	55.8
−300	2	56.8	57.1	55.5	56.8	62.1	54.9	58.8	60.4	61.5	58.2
−150	3	56.3	59.4	55.6	59.4	64.9	57.3	59.9	62.6	60.3	59.5
−75	4	57	61.3	56.2	61.9	66.4	57.5	60.8	63.2	61.9	60.7
0	5	55.2	58.7	53.3	59.9	63.3	54.9	59.5	61.8	60.6	58.6
50	6	50.6	54.7	49.6	56	58.5	51.7	56.9	57.4	56	54.6
100	7	48.3	51.6	47.4	51.4	54.4	54.4	50.5	53.6	52.3	51.5
	IVT, kg m−1 s−1
−450	1	199	269	253	268	372	373	460	422	305	327
−300	2	221	286	170	379	328	381	506	537	342	349
−150	3	203	182	201	347	288	223	366	470	251	279
−75	4	174	221	250	413	424	240	422	544	275	326
0	5	77	114	178	233	264	76	215	223	118	165
50	6	84	76	124	140	114	43	63	64	31	83
100	7	40	72	105	86	45	91	98	89	58	76
	P−E, mm h−1
−375	1 → 2	−0.5	−0.4	2.0	−2.7	1.1	−0.2	−1.1	−2.8	−0.9	−0.6
−225	2 → 3	0.4	2.5	−0.7	0.8	0.9	3.8	3.4	1.6	2.2	1.6
−112.5	3 → 4	1.4	−1.9	−2.4	−3.2	−6.5	−0.8	−2.7	−3.6	−1.1	−2.3
−37.5	4 → 5	4.7	5.1	3.5	8.6	7.7	7.8	10.0	15.4	7.5	7.8
25	5 → 6	−0.6	2.7	3.9	6.7	10.8	2.4	10.9	11.5	6.2	6.1
75	6 → 7	3.2	0.3	1.3	3.9	4.9	−3.4	−2.5	−1.8	−1.9	0.4

Note(s): Coordinates of the 7 points: 0.23° N, 100.49° E; 0.02° S, 100.04° E; 0.20° S, 99.82° E; 0.36° S, 99.17° E; 0.67° S, 98.63° E; 1.44° S, 97.47° E and 1.44° S, 95.31° E, respectively.

Figure 11. (A). Hourly estimated P−E for November 2025 in Figure 9; (B). Existing maps of average 5-day rainfall accumulation over and around Sumatra in TRMM satellite-derived estimates in reference [52] with the transect of Figure 10 as red insert.

5. How Does Land Cover Influence Flow Buffering and Turbulent Flows After Peak Rainfall Events (RQ2)?

Beyond a ‘deforestation’ discourse, we need to understand the buffer factor of the land cover that replaced natural forest, whether or not this is still within the ‘forest’ category of land use categories. Converting natural forests to other tree-based systems alters the hydrological services provided by the cover, including canopy roughness, which may influence atmospheric moisture transport and rainfall patterns, and infiltration regulation that influences surface runoff. Therefore, the role of forests in flood hazard management should be considered holistically, not only by assessing each vegetation, soil and rainfall individually but also by the interaction of these three components.

The processes listed in Figure 3 indicate the main aspects of land cover/land use that need to be considered for any land cover type:

• Atmospheric roughness (heterogeneity of tree heights), typically high for mixed-age and mixed-species stands and boundary plantings, low for even-aged monocultures;
• Rainfall nuclei influencing the critical temperature for raindrop formation;
• Leaf Area Index and its phenology or seasonal pattern;
• Surface litter layer due to varied litterfall rates and qualities;
• Woody roots that explore subsoil;
• Macroporosity as generated by root turnover and ‘soil engineers’ among the biota supported.

Flood hazard regulation needs to be viewed more broadly, not only by assessing each vegetation, soil and rainfall individually but also by the interaction of these three components. Merten et al. [55] attributed changes in local flooding regimes of the Tembesi river in Jambi (Sumatra, Indonesia) (after accounting for specifics of the rainfall pattern) to an increase in surface runoff due to soil compaction after the conversion of forests and ‘jungle rubber’ to monoculture plantations (rubber, oil palm) and to the increasing encroachment and conversion of naturally vegetated wetlands.

Specifically in forested areas, there is a risk for increased flood risks where logjams or debris dams build up and initially delay and, upon breakage, increase peak flows. As deep landslides are part of a natural forest dynamic, the sudden blockage of streams can occur in natural forests, but the probability becomes much higher where ongoing logging operations increased the presence of logs in the landscape. A difference between the two causes of debris dams is in the presence of trees with intact root systems, versus cut stems.

6. Exposure to Flood Hazards in Various Parts of Sumatra (RQ3)?

From the rich sources of data [1] on actual damage across the various parts of Sumatra affected, it is clear exposure has been hard to avoid in several settings:

• Coastal zones along the west coast where floods happened even though W-facing mountain slopes had generally protected forest cover; yet, tree fall and landslides caused increased risk along the river channels, some of which are densely populated.
• Inland areas, where bridges washed out and road access was blocked, sometimes for weeks before road access could be re-established.
• Floodplains and large irrigate agriculture schemes along the N and E coasts of Sumatra, where rainfall intensity substantially exceeded water buffering options, especially where peat and drainage-cased subsidence had already increased groundwater tables. Turbulent flow of rivers had mobilized large volumes of soil particles that sedimented on crop fields and destroyed the existing crops.

(North)-east coast floods occurred due to insufficient water storage capacity, aggravated by peat subsidence; local rainfall already exceeded storage [56], with additional river flow aggravating the damage (and adding siltation problems). While the Senyar Flood is a rare event, it still has the potential to occur due to the confluence of certain atmospheric and biophysical conditions. Therefore, the water storage capacity achieved under “normal” flood conditions may not be applicable during ‘new normal climate’ that lead to a ‘new’ (extreme) flood condition. The government and community may have considered spatial planning for residential and productive areas within “normal” flood-prone areas. However, the government and community still need to develop mitigation scenarios for areas vulnerable to “extreme” flooding when such events occur again in the future.

Some recommended mitigation scenarios for the government and community to implement in facing the “new normal climate” include the following:

• The drainage system in densely populated settlement and industrial areas along roads is designed for rapid water disposal, increasing flooding problems downstream. Redesign, re-construction and maintenance need to combine local surface water storage areas for groundwater recharge and excess water disposal without undue damage downstream. Managed resettlement of flood-prone people must be sensitive to their needs and rights.
• Water management and storage plans in irrigated and drained lowlands need to adjust to ‘new normal’ climates.
• Existing drainage standards for plantations need to be re-assessed for their external environmental impacts as well as production levels. Instead of further expansion of plantation areas, sustainable intensification (closing yield gaps, currently around 50% in oil palm, for example) can allow for increased production (say from 50 to 80% of potential yields). Mixed gardens with perennial crops (‘kebun lindung’) offer lessons in managed trade-offs, with permanent protective soil cover.
• The legal obligations to maintain the protective functionality of riparian zones need to be enforced and monitored. More space for rivers upstream protects downstream areas from flood peaks.
• People in remote villages with a single road as access need to store essentials and be prepared that transport can be interrupted.
• Understanding how debris dams originate and where and how landslides can start and be filtered in remaining primary and secondary forests on slopes (>15%) needs to inform management.
• Land suitability and downstream vulnerabilities need to be an explicit part of zoning and (re)new(ed) concession permits.

7. What Lessons Can Be Learned for Building Back Better (RQ4)?

There is a long tradition in classifying watershed degradation’ based on ‘forest cover’ with limited use of hydrological data (the popular maximum/minimum flow metric is not applicable where streams are intermittent and the minimum is zero). Aligned with the history of land use, the issues that arose and the solutions created, a patchwork of concepts deals with ‘land degradation’/’restoration’, each with its own idea of relevant metrics (what can be measured to establish priorities and monitor progress). Table 3 gives an overview.

Table 3. Flood risk problems and solutions across hazard, exposure, and vulnerability (modified from [4]).

Category	Problem	Solution/Recommendation
Hazard (from Sumatra level to site-specific effects)	Increasing occurrence of compound extreme events (e.g., Cyclone Senyar–Ditwa interaction), amplified by climate anomalies (ENSO, IOD)	Strengthen regional weather monitoring, forecasting and early warning systems. Integrate atmospheric and land-surface hydrology
	Climate change disrupts historical baselines, making events harder to predict	Shift from reliance on historical statistics to process-based planning under a shifting climate ‘normal’
	Lack of recognition of land-atmosphere feedback (e.g., forests and trees influencing rainfall and cooling)	Explicitly include forest-water-climate linkages in adaptation policy and financing mechanisms
	Decline in landscape buffering due to deforestation, soil compaction, conversion to plantations	Promote hydrologically functional landscapes that can include agroforestry, wetlands, terracing and infiltration-friendly land uses
	Clear-felling and logging debris aggravates downstream impacts through ‘logjams’	Enforce environmental safeguards; strengthen monitoring of forest concessions; support evidence-based zoning (e.g., of protection forest categories)
Exposure (in flooded locations)	Settlements and infrastructure located in high-risk zones due to poor or non-implemented spatial planning	Improve risk-informed land use planning, enforce zoning and support community-based early response mechanisms
Vulnerability (when exposed to high water levels)	Reclassification of locally adaptive systems (e.g., protective gardens or ’kebun lindung’) as ‘state forest’ reduced resilience	Recognize and integrate locally managed, multifunctional systems in formal land use and adaptation planning
	Weak institutional coordination across forest, water (public works, agriculture) and disaster sectors	Develop shared diagnostic and integrative governance frameworks at multiple scales
	Public narratives oversimplify causes (e.g., the deforestation leads to floods slogan) and distract from systemic solutions	Encourage multi-perspective dialogue and Science-Policy communication; counter scapegoating with systemic diagnosis
	Climate adaptation policies are often disconnected from flood realities-on-the-ground	Reconcile UNFCCC adaptation frameworks with hazard-exposure-vulnerability logic from the disaster risk reduction traditions
	Unequal access to resources and limited adaptive capacity in affected communities	Invest in human and institutional capacity, e.g., guided by the Inner-Development-Goals

The simplest way to avoid damage may be to ensure that nobody is exposed. If at the time of the Dec 2004 Tsunami, all mangrove areas would have been intact and not inhabited, nobody would have drowned there. The strongest ‘explanations’ for flood damage are human population density. As this is negatively related to forest cover, it may seem that forest cover actively protects people—maybe it does in a given (or even many) context(s)—but for a clean analysis of disaster ‘hazard’, it needs to be considered from ‘exposure’, while both combine to ‘explain’ (human) vulnerability (Table 3). Where reduced flow buffering can reduce local and increase downstream flooding hazard, there is some unavoidable ambiguity in how ‘exposure to hazards’ interacts with ‘hazards’ as such at the landscape or watershed scale.

Vulnerability (likelihood of victims) is the result of ‘hazard’ (extreme events to occur) and ‘exposure’ (being at wrong time at wrong place). Exposure is easier to control than hazard, and measures to reduce human vulnerability to floods include the following:

○

Not (re)building houses in the likely course of flash floods in the local river systems;

○

Not (re)building houses at places exposed to landslide risk;

○

Not (re)building cities on ‘flood plains’, even though floodplains may have fertile soil and are close to rivers as economic access options;

○

If floodplains have still been developed into settlement areas, select the highest places and/or protect selected areas using dykes and drainage canals;

○

Ensure that bridge design and construction are accompanied by appropriate risk analysis;

○

Create early-warning systems that lead to the temporary evacuation of at-risk locations;

○

Have risk awareness built into all aspects of water management, as all parts are connected.

8. Discussion

Runoff genesis in mountain catchments can be modified by protective forests in multiple ways. Beyond the commonly discussed ‘downstream’ highland-to-lowland effects, our analysis suggests that ‘downwind’ impacts from lowlands to highlands may be involved as well. For the specific cyclone-induced floods, the spatial distribution of damage appears to be incompletely understood. Recorded windspeeds in space and time after cyclone landfalls have often been described as exponential decay models, with regional variation in decay rates that can be attributed to multiple factors [57]. When cyclone (hurricane) Maria crossed Puerto Rico with windspeeds as high as 250 km h⁻¹, it resulted in widespread damage but also in loss of weather station data. After the event, spatial patterns in tree breakage could be used to model the distribution of cyclone windspeed when ground readings were sparse [58]. A case study for the east coast of India found that mangrove tree cover influenced a windspeed reduction after cyclone landfall but identified a need for further quantification [59].

Complex causation of rainfall with at least ten (historical) concepts may coexist in contemporary public discussions [60]. Our hypothesized shift from ‘vertical’ (temperature-based) to horizontal (inflow exceeds outflow) concepts of ‘orographic’ rainfall is aligned with how windbreaks work (effect (topographically and temporally) before the actual barrier but deserve further analysis and communication efforts. The cyclone landfall can add arguments to a long-standing debate:

• The ‘Biotic Pump’ theory [24,61] explains positive effects of forest on rainfall, not only based on ‘short cycle’ recharge of atmospheric fluxes by evapotranspiration but also by suggesting forests (by evaporative cooling) induce wind that transports the moisture inland.
• The ‘Prevailing Winds’ alternative [30] to the Biotic Pump theory accepts wind as part of latitude-dependent atmospheric circulation systems but emphasizes quantification of atmospheric moisture balances.
• The cyclone landfall literature suggests that tree cover reduces (rather than increases as the Biotic Pump theory assumes) windspeed. If land–ocean interface changes in windspeed can indeed be used to predict ‘flux conserving’ rainfall, it is the reductions in windspeed slowing down winds that affect rainfall (rather than increase in windspeed).

We have focused on publicly available spatially and temporally explicit representations of atmospheric fluxes that have a complex relationship with various types of primary meteorological measurements and empirical models—a full uncertainty analysis is currently beyond our capabilities. As emphasized by part of the literature, process-level interactions between temperature, air pressure and the condensation or evaporation of water (within the TPW category) include positive feedback loops that can generate strongly non-linear responses.

Reduced atmospheric roughness due to forest conversion in the coastal zone between the Batang Toru watershed and the ocean [62,63] may have increased the rainfall over the mountain range and contributed to the landslides documented in [54]. A change in roughness may have little effect on the total amount of rainfall received by a coastal watershed but can modify the spatial distribution. In this situation, it is lowland forests that protect forests higher up the slopes, rather than the other way around.

The relationship between deforestation and increased flooding has been much debated. It may help to distinguish between effects on hazard (modified rainfall, modified river flow), on exposure (e.g., previously forested locations become settlements, as happens where urban mangroves are converted) and effects on vulnerability as such. The deforestation → flood discourse may involve the following steps:

• Flood hazards appear to increase in frequency/duration/intensity (at least when a recent event sparks interest);
• There are plausible causal links with increased weather variability due to global climate change;
• Forests continue to be converted and/or degraded due to (legalized?) large-scale operations and/or (illegal?) small farmers [63];
• Urban areas keep expanding, reducing flood tolerance unless engineering interventions are effective;
• Unless exposure is managed and reduced, increased hazards contribute to increased vulnerability, especially for those with low tolerance (‘already vulnerable’);
• It matches policy agendas if A can be linked to B, to C or both;
• Realistic damage minimization policies embrace points D, E and F.

9. Conclusions

The following findings are supported by this review:

• Runoff genesis in mountain catchments responds to the way an incoming atmospheric moisture flux at the ocean–land interface leads to (peak) precipitation (topographically and temporally), before ‘downstream’ flow processes start.
• The rare near-equatorial Cyclone Senyar depended on repeatable coincidences that may become more frequent in a changing climate. Statistically observed frequencies are no longer a safe basis for risk assessment.
• The substantial increase in remotely sensed data is not yet matched by ‘theories of place’ and ‘theories of change’ that can effectively inform current actions. Desirable future empirical work includes the dissection of the terrain and vegetation interaction in atmospheric roughness effects.

Hypotheses generated by the conceptual analysis:

• Our analysis suggests that a slowdown in atmospheric flows, along with convergence of streamlines, can induce ‘congestion’ with precipitation as consequence when the temperature-dependent maximum TPW value has been reached (as vertical movement would be incorporated in the TPW data).
• Two consequences of the ‘congestion’ perspective are that any forest influence on windspeed reduction may directly link to extreme rainfall and that orographic rain may involve (horizontal) windbreak effects before cooling of (vertically) uplifted air masses. Further scrutiny of these hypotheses is needed.

Recommendations for research and risk reduction:

• Further work is also needed on quantification of flow buffering, plantation drainage systems and ‘space for the river’ concepts can reduce exposure of lowland flood-prone areas on former floodplains. Lack of upstream water buffering has likely been a cause of flood damage to exposed people and homes, adding to location-specific rainfall, but the relative proportion will have varied between locations in ways that deserve to be further unraveled.
• Future adaptation and disaster avoidance efforts should seek to balance the reduction in human exposure through effective land use planning and efforts to reduce hazard by restoring and managing vegetation cover and drainage systems.
• As part of new interdisciplinary efforts by Indonesian scientists to analyse the various backgrounds of the substantial damage in the aftermath of the Senyar Cyclone in Sumatra, we hope that there will be space for a diversity of views and perspectives, rather than a rapid choice for a single simplified narrative (such as ‘tree planting’) that will dominate recovery actions.