Vishal Batchu

Vishal Batchu

Auto-generated profile, based on publicly available information

Vishal Batchu

Google Research, Google

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 October 2023 in Journal of Hydrometeorology
Verified
Journal Article
Closed access
PDF file
VBVishal Batchu
Vishal Batchu
GNGrey Nearing
Grey Nearing
VGVarun Gulshan
Varun Gulshan
AHA. J. Henry
A. J. Henry
CLClose
Close
We develop a deep learning–based convolutional-regression model that estimates the volumetric soil moisture content in the top ∼5 cm of soil. Input predictors include Sentinel-1 (active radar) and Sentinel-2 (multispectral imagery), as well as geophysical variables from SoilGrids and modeled soil moisture fields from SMAP and GLDAS. The model was trained and evaluated on data from ∼1000 in situ sensors globally over the period 2015–21 and obtained an average per-sensor correlation of 0.707 and ubRMSE of 0.055 m3 m−3, and it can be used to produce a soil moisture map at a nominal 320-m resolution. These results are benchmarked against 14 other soil moisture evaluation research works at different locations, and an ablation study was used to identify important predictors. Significance Statement: Soil moisture is a key variable in various agriculture and water management systems. Accurate and high-resolution estimates of soil moisture have multiple downstream benefits such as reduced water wastage by better understanding and managing the consumption of water, utilizing smarter irrigation methods and effective canal water management. We develop a deep learning–based model that estimates the volumetric soil moisture content in the top ∼5 cm of soil at a nominal 320-m resolution. Our results demonstrate that machine learning is a useful tool for fusing different modalities with ease, while producing high-resolution models that are not location specific. Future work could explore the possibility of using temporal input sources to further improve model performance.

1 September 2026 in Proceedings of the National Academy of Sciences
Verified
Journal Article
Open access
PDF file
VBVishal V. Batchu
Vishal V. Batchu
MCMichelangelo Conserva
Michelangelo Conserva
AWAlex Wilson
Alex Wilson
AMAnna M. Michalak
Anna M. Michalak
VGVarun Gulshan
Varun Gulshan
CAChristopher V. Arsdale
Christopher V. Arsdale
Anthropogenic methane (CH 4 ) point sources are critical drivers of near-term climate forcing, safety hazards, and system-inefficiencies. Space-based imaging spectroscopy is an emerging tool for identifying emissions globally, but existing approaches largely rely on manual plume identification. Here, we present the Methane Analysis and Plume Localization with Earth Surface Mineral Dust Source Investigation (MAPL-EMIT) model, an end-to-end vision transformer framework that leverages the complete radiance spectrum from the EMIT instrument to jointly retrieve methane enhancements across all pixels within a scene. This approach brings together spectral information with spatial context to significantly lower detection limits. MAPL-EMIT simultaneously supports enhancement quantification, plume delineation, and source localization, even for overlapping plumes. The model was trained on 3.6 million physics-based synthetic plumes injected into global EMIT radiance data. Synthetic evaluation confirms the model’s ability to identify plumes with high recall and precision and to capture weaker plumes relative to existing matched-filter approaches. On real-world benchmarks, MAPL-EMIT captures 84% of known hand-annotated NASA EMIT-L2B plume complexes across a test set of 1,084 EMIT granules, while capturing roughly 1.5 times as many plausible plumes compared to human analysts. Further verification against coincident airborne data, top-emitting landfills, and controlled release experiments confirms the model’s ability to identify previously uncaptured sources. By incorporating model-generated metrics such as spectral fit scores and estimated noise levels, the framework can limit false-positives. Overall, MAPL-EMIT enables high-throughput implementation on the full EMIT data catalog, shifting methane monitoring to a rapid, scalable paradigm for global plume mapping at the facility scale.

16 October 2023 in arXiv
Verified
Preprint
Open access
PDF file
VBVishal Batchu
Vishal Batchu
GNGrey Nearing
Grey Nearing
VGVarun Gulshan
Varun Gulshan
We develop a deep learning based convolutional-regression model that estimates the volumetric soil moisture content in the top ~5 cm of soil. Input predictors include Sentinel-1 (active radar), Sentinel-2 (optical imagery), and SMAP (passive radar) as well as geophysical variables from SoilGrids and modelled soil moisture fields from GLDAS. The model was trained and evaluated on data from ~1300 in-situ sensors globally over the period 2015 - 2021 and obtained an average per-sensor correlation of 0.727 and ubRMSE of 0.054, and can be used to produce a soil moisture map at a nominal 320m resolution. These results are benchmarked against 13 other soil moisture works at different locations, and an ablation study was used to identify important predictors.

22 June 2026 in Nature Communications
Verified
Journal Article
Open access
DFDavid Fork
David Fork
EWElizabeth Wesley
Elizabeth Wesley
SBSalil Banerjee
Salil Banerjee
VBVishal Batchu
Vishal Batchu
ACAniruddh Chennapragada
Aniruddh Chennapragada
LHLucy R. Hutyra
Lucy R. Hutyra
Extreme urban heat poses a growing threat to global health, yet a critical cooling intervention-reflective cool roofs-lacks the high-resolution data needed for building-level planning. While satellite-based albedo estimates are globally available, their coarse resolution cannot resolve individual rooftops, limiting targeted climate action. Here we show that fusing freely available 10-m Sentinel-2 imagery with high-resolution satellite data enables urban albedo mapping at 30-cm resolution. This approach combines the radiometric accuracy and global coverage of Sentinel-2 with the spatial detail of commercial imagery to produce actionable urban datasets. Validated against airborne hyperspectral measurements (root mean square error = 0.04), the method resolves albedo at individual building footprints across diverse urban environments. Applied to 12 global cities, our analysis shows that prioritizing large-footprint buildings for cool roof retrofits can reduce citywide temperatures by up to 0.5 °C. This work validates Sentinel-2 for city-scale albedo estimation and enables globally deployable, building-scale rooftop albedo mapping to support targeted cool roof interventions.

8 August 2026
Verified
Conference Paper
Open access
RDRadhika Dua
Radhika Dua
AAAditi Agarwal
Aditi Agarwal
AJAishwarya Jayagopal
Aishwarya Jayagopal
DSDepanshu Sani
Depanshu Sani
AWAlex Wilson
Alex Wilson
VRVaibhav Rajan
Vaibhav Rajan

11 November 2026 in EarthArXiv
Verified
Preprint
Open access
PDF file
VBVishal Batchu
Vishal Batchu
MNMaxim Neumann
Maxim Neumann
TLTommy Li
Tommy Li
BABrett Allen
Brett Allen
JZJie Zhang
Jie Zhang
CAChristopher Van Arsdale
Christopher Van Arsdale
Existing open global digital elevation models (DEMs) are limited to 30 m resolution, blur fine-scale urban and canopy structure, and rely on satellite missions flown between 2000 and 2015. We present Google's Global Elevation Models at 15 m (GEM-15), to our knowledge the first openly available near-global multi-layer elevation dataset at 15 m grid spacing. GEM-15 aggregates native 0.5 m optical stereo reconstructions acquired through 2024 (63.2% of mapped tiles primarily from 2022–2024 30 cm Pléiades Neo imagery) and provides four times the spatial sampling density of 30 m baselines, more current representation of the Earth's surface, and higher vertical accuracy over complex canopies, buildings, and steep relief. Alongside the primary 15 m digital surface model (DSM), GEM-15 includes within-cell elevation percentiles (P5, P50, P95), local elevation standard deviation (DSM σz), and a co-registered 15 m bare-earth digital terrain model (DTM) to capture within-cell 3D structure. We evaluate GEM-15 alongside the open 30 m DEMs Copernicus GLO-30, ALOS AW3D30, NASADEM and, for the terrain model, FABDEM against spaceborne GEDI lidar, 12 airborne and drone lidar benchmarks, and 4,071 GNSS CORS stations. Across the lidar benchmarks, the GEM-15 DSM has a 12.3% lower mean tile-debiased root-mean-square error (DRMSE) than Copernicus GLO-30 (3.37 m vs. 3.85 m), rising to 13.7% (3.69 m vs. 4.27 m) for the seven benchmarks with acquisition midpoints from 2021 onward; the margin increases with the reference survey date (Spearman ρ = 0.87 across the 10 non-rainforest campaigns) and reaches 27% for the 2024–2025 New Zealand survey. The co-registered DTM is comparable in accuracy to FABDEM in open terrain and less accurate beneath closed canopy. In repeat surveys at US NEON sites, the 30 m DEMs diverge increasingly from recent lidar, whereas GEM-15 has the lowest DRMSE in every year. The dataset is available under a CC BY 4.0 license at https://developers.google.com/earth-engine/datasets/catalog/projects_climate-and-sustainability_assets_gem-15_v1.