Sedimentary Records from Human-Made Talavs Reveal Climate Risks in Semi-Arid Watersheds of India

Introduction

• South Asia's arid and semi-arid regions (SARs) are climate hotspots vulnerable to anthropogenic climate change impacts.
• Watersheds in peninsular India are particularly threatened, exhibiting tight human-environment coupling sensitive to climatic extremes.
• Historical examples: Famines caused by minor rainfall fluctuations due to water-intensive cash crop expansion (cotton, sugarcane) during British rule (18th-20th centuries).
• Recent consequences of anthropogenic climate change: Crop failures, rural-urban migration, and farmer suicides, increasing economic stress and social instability (late 20th century).
• Need for improved climate risk understanding in diverse climatic zones and microclimatic habitats.
• High-resolution environmental records in sediments of surface water bodies, especially human-made reservoirs, offer valuable insights.
• This study reconstructs catchment-level environmental history from sediment cores of talavs (human-made water reservoirs) in the Bhima watershed, a climate hotspot in peninsular India.

What are Talavs?

• Talavs are surface water bodies created by damming tributaries and distributaries of seasonally rain-fed river channels in arid and semi-arid South Asia.
• Typical talav size: Area ~ 2-5 $km^2$, catchment area ~ 30-100 $km^2$.
• Historical significance: Mentioned in archeological studies, historical texts, and contemporary policy literature.
• Commissioned by Mughal rulers (12th century) and commonly constructed by British administrators (late 19th century) and the Indian government after 1947.
• Purpose: Initially for providing water during dry spells, then for commercial agriculture of water-intensive crops, and recently for hydroelectricity generation.
• Talavs trap and store sediments, enabling reconstruction of catchment environmental history and providing insights into climate risks in watersheds.
• This study analyzes sediment cores from Matwali (MAWL) and Indira Bazar (IDBZ) talavs in the Bhima watershed, representing Madhya Maharashtra and Marathwada climate divisions respectively.

Study Area

• The study area is situated along the rain-fed semi-arid Bhima watershed on the Deccan trap basalts, lying in the rain shadow of the Western Ghats and Balaghat ranges.
• The Bhima watershed covers two Indian Meteorological Division (IMD) climatological divisions:
  • Wetter Madhya Maharashtra in the northwest.
  • Drier Marathwada in the southeast.
• Rainfall is delivered by the Indian southwest monsoon (ISM), making surface water bodies sensitive to ISM variations.
• Two talavs were studied:
  • Matwali talav (MAWL): Located in Beed district (near Balaghat ranges) in the Marathwada climate division (18.797893° N, 75.290229° E).
  • Indira Bazar talav (IDBZ): Located near Pune in the upper Bhima basin in the Madhya Maharashtra climate division (18.896760° N, 73.903462° E).
• Both talavs were created by constructing embankments on distributaries of the Bhima River, making them suitable for recording erosional patterns in their catchments.

Materials and Methods

• Geochemical proxies in semi-arid regions are sensitive to hydroclimate variations and associated environmental changes.
• This study uses geochemical proxies in sedimentary cores from MAWL and IDBZ talavs to reconstruct erosional patterns associated with hydroclimate variability.
• A multi-disciplinary approach was used, combining:
  • Satellite remote sensing image analysis.
  • Seismic data acquisition, processing, and interpretation.
  • Sedimentary coring.
  • Downcore sedimentological and geochemical measurements.
  • Radiometric dating of ~1-meter cores from MAWL and IDBZ.
• The goal was to constrain erosion and sedimentation patterns and assess climate risks across the Bhima watershed.

Satellite Image Analysis

• Used USGS Landsat 1-3 and Landsat 5 Multispectral Scanner and Thematic Mapper terrain-corrected satellite images (1970 onwards).
• Created standard false color composites (near-infrared, red, and green bands), Normalized Difference Vegetation Index (NDVI), and Normalized Difference Water Index (NDWI) to identify vegetated areas and water bodies.
• Created a time-series of land-cover changes in the two catchments from the mid-1970s to the present.
• Employed Shuttle Radar Topography Mission (SRTM) digital elevation data (90-meter spatial resolution) for topographic variations in slope analysis.

Seismic Profiling

• Conducted two-dimensional (2D) seismic profiling to image the subsurface of the lakes.
• Seismic source: GeoPulse transducer (5430 A) transmitting a chirp signal in the 2–12 kHz frequency bandwidth.
• Receiver: Recorded the returns.
• Signal penetration: Up to 2–3 meters in the sand/silt environment.
• Water depths: Ranging from 2 to 5 meters.
• Navigation: Hemisphere R100 DGPS integrated with the CodaOctopus DA4G system.
• Seismic lines were collected along a predefined grid and converted into a depth profile using a water velocity of 1500 m/s.
• Software used: Geosurvey processing and S&P Global Kingdom.
• Analyzed seismic facies and defined seismic stratigraphic units using reflections.
• Determined coring locations and correlated seismic stratigraphy with coring units.

Sediment Core Extraction and Sampling

• Retrieved sediment cores from in-filled gullies in the deepest parts of IDBZ and MAWL talavs in May 2018.
• Coring techniques: Vibra-coring, gravity coring, and push-coring.
• IDBZ talav: Push-coring method used because the lake water was drained.
• Longest core retrieved: ~1.2 meters from MAWL, 78.5 cm from IDBZ.
• Cores were wrapped, labeled, and shipped to the National Lacustrine Core Facility at the University of Minnesota.
• Cores were longitudinally split, described, scanned for magnetic susceptibility, and photographed at high-resolution.
• Sediment cores were sub-sampled at 1-cm intervals and refrigerated.
• Cores remained archived at the National Lacustrine Core Facility.

Radiometric Dating

• Used 14C and 137Cs dating methods to date MAWL and IDBZ sediments.
• Radiocarbon dating:
  • Measured 14C concentrations in ~0.025 mg of separated terrestrial organic material (wood, charcoal, seeds).
  • Samples were treated with acid-base-acid (1 N HCl and 1 N NaOH, 75°C) prior to combustion.
• 137Cs dating:
  • Weighed ~3 g of dried sediment from seven unique samples (two from MAWL, five from IDBZ).
  • Homogenized and analyzed for 137Cs concentration at the St. Croix Watershed Research Station.
  • Used an Ortec-EGG High-Purity, Germanium Crystal Well, Photon Detector coupled to a Digital Gamma-Ray Spectrometer.
  • 137Cs activity was measured using GammaVision software.

X-Ray Fluorescence (XRF)

• Cores were scanned using an Itrax Core scanner at the Large Lakes Observatory, University of Minnesota-Duluth.
• Surface of each core was gently scraped with a glass slide prior to scanning.
• Used a Cr source tube run at 30 kV and 55 mA, with a 5-mm resolution and a 15-second dwell time.
• Raw data were reprocessed to optimize peak fitting using QSpec 8.6.0 software.

Particle Size Analysis (PSA)

• PSA for sediment samples from MAWL and IDBZ cores was carried out using a laser particle size analyzer (Malvern mastersizer 3000) coupled to a Hydro EV wet dispersion unit at the Indian Institute of Science Education Research (IISER)-Mohali.
• Samples were pretreated with 10 ml $H2O2$ (30 %) to remove organic matter and then washed with deionized water.
• 1 N HCl acid was added to remove carbonates, followed by washing with deionized water.
• The solid residue was kept in ~20 ml double-distilled water.
• Samples were ultra-sonicated for 15 minutes prior to analyzing.
• Grain size classes ranged from 0.02 to 2000 µm, with an analytical error < 1 %.

Thin Sections

• Optically examined 26 unique smear slides, prepared from core sections of IDBZ and MAWL at the Lacustrine Facility in Minnesota, at the Indian Institute of Technology, Kharagpur.
• Used a Leica Orthoplan Pol microscope under plane polarized and crossed polarized lights.

X-Ray Diffraction (XRD)

• A total of nine unique samples from both sites were scanned on the Rigaku Miniflex at the Continental Scientific Drilling Facility, University of Minnesota.
• Smear samples were prepared on glass slides and run using a cobalt anode with a scan range of 5–65° 2θ, step width of 0.02° 2θ, and scan speed of 1° 2θ min− 1.
• Mineral phases were identified using Profex software, including background subtraction, peak identification, refinement, and matching with XRD patterns of reference compounds.

Results

• Results are summarized in Figs. 3–6 and the Supplemental Information (SI) (Tables SI 3–7 and Figs. SI 2–8).

Embankment Structure

• The modern embankment structure at MAWL was constructed between 1989 and 1990.
• The embankment at IDBZ was completed between 1973 and 1975.
• MAWL talav is prone to complete desiccation, whereas IDBZ talav was not observed to dry up completely.
• Human activity (agriculture and grazing) was confined to the edges of the water bodies.

Talav Sedimentary Sequence

• In both talavs, a laminar sedimentary sequence (Unit I) overlies a coarser sedimentary deposit (Unit II).
• Seismic surveys show reflection patterns that serve as unique geophysical representations of the reservoir's sedimentary system.
• Sediment thickness varies in Unit I at MAWL, with more sediment in channels (~80 cm thick) than in surrounding regions (45–50 cm).

Characteristics of Deposits

• Sediments deposited in the talavs represent erosion from basement rocks within the catchment (Deccan traps) and show laminated deposition patterns.
• Coarse-grained, poorly-sorted deposits (plugs) are present only at the MAWL site.
• Laminated structure within the MAWL core is dominated by clinopyroxene (Cpx), plagioclase feldspar (Plg), and iImenite (Ilm), typical of Deccan basalts.
• The poorly-sorted sandy interval (47–62 cm) at the MAWL site is dominated by feldspars, with less clinopyroxene (Cpx).
• The sediment core collected from the IDBZ talav, predominantly consists of dark brown massive silt interbedded with distinct pale brown laminae of well-sorted silty clay with low organic carbon and organic carbonate content.

Age Determination

• Radiometric dating of the lowest part of the cores demonstrates that sediments in both locations are modern (deposited during the 20th and 21st centuries).
• Results of 137Cs dating show only trace amounts of Cesium in the MAWL core from 69 to 80 cm, confirming their age as at least pre-1950.
• Low 137Cs activity from 52 to 82 cm stratigraphic depth in the IDBZ core, without a clear 1963 peak.

Downcore Grain Size and XRF Data

• Downcore grain-size analyses and XRF scanning reveal uniformly varying elemental ratios of common bedrock elements associated with laminar deposition in Unit I of both MAWL and IDBZ cores.
• Median grain size at MAWL is ~50 microns throughout the core except in the poorly-sorted interval (47–62 cm core depth), where it is ~150 microns.
• The average grain size in the IDBZ core is smaller ~5 microns.
• Principal component analysis of XRF data shows clear distinction between authigenic and detrital fractions of the core sediments.
• Downcore variation in grain size data and hydroclimate-sensitive proxy of Ti/Al ratios are analysed in relation to risks arising from hydroclimate variability and extremes.

Discussion

• Sedimentation in talavs is influenced by hydroclimate variability and anthropogenic land-use changes.
• Catchments are likely to reveal stronger erosion during Indian summer Monsson
• The length of time represented by the sediment cores is determined and then related against Hydroclimate Variability.