arXiv subjects

Rajat Masiwal

Publications and source records attributed to Rajat Masiwal.

3 recordsLinked to original sources

arXiv2026

Designing probabilistic AI monsoon forecasts to inform agricultural decision-making

Hundreds of millions of farmers make high-stakes decisions under uncertainty about future weather. Forecasts can inform these decisions, but available choices and their risks and benefits vary between farmers. We introduce a decision-theory framework for designing useful forecasts in settings where the forecaster cannot prescribe optimal actions because farmers' circumstances are heterogeneous. We apply this framework to the case of seasonal onset of monsoon rains, a key date for planting decisions and agricultural investments in many tropical countries. We develop a system for tailoring forecasts to the requirements of this framework by blending systematically benchmarked artificial intelligence (AI) weather prediction models with a new "evolving farmer expectations" statistical model. This statistical model applies Bayesian inference to historical observations to predict time-varying probabilities of first-occurrence events throughout a season. The blended system yields more skillful Indian monsoon forecasts at longer lead times than its components or any multi-model average. In 2025, this system was deployed operationally in a government-led program that delivered subseasonal monsoon onset forecasts to 38 million Indian farmers, skillfully predicting that year's early-summer anomalous dry period. This decision-theory framework and blending system offer a pathway for developing climate adaptation tools for large vulnerable populations around the world.

cs.LG↗

arXiv2026

Emergence of an Advective Boundary Layer in Monsoon Cross-Equatorial Flow: Scaling, Dynamics, and Idealized Models

The conventional Ekman model of the tropical boundary layer neglects nonlinear momentum advection and breaks down near the equator, where Coriolis effects are weak. During South Asian monsoon onset, we identify a dynamical regime transition to an advective boundary layer (ABL). Reanalysis links this transition to a shift in the zonal momentum balance from frictional to meridional-advection control as cross-equatorial flow intensifies, accompanied by increasing local Rossby number and vanishing absolute vorticity, signaling the breakdown of Ekman balance. A scaling analysis shows that this transition occurs when the meridional length scales of geopotential and zonal wind contract such that their product approaches $ϕ/f^2$. In the resulting ABL regime, kinetic energy is governed by a balance between its generation and advection, yielding a linear diagnostic relation between meridional geopotential gradient and meridional wind. A simple theoretical model predicts that the sensitivity of this relation is controlled by an advective timescale that equals the inertial timescale ($1/f$) at the transition latitude, where zonal and meridional wind speeds become comparable. Testing this framework in idealized aquaplanet experiments confirms that stronger cross-equatorial pressure gradients and slower planetary rotation rates amplify advective effects and shift the transition latitude poleward. Across experiments, the sensitivity of meridional winds to the geopotential gradient remains tightly linked to $1/f$ at the transition latitude. Together, these results establish the ABL as a distinct dynamical regime, with important implications for monsoon onset, intraseasonal variability, and the representation of tropical boundary layer processes in climate models.

physics.ao-ph↗

arXiv2026

Decision-oriented benchmarking to transform AI weather forecast access: Application to the Indian monsoon

Artificial intelligence weather prediction (AIWP) models now often outperform traditional physics-based models on common metrics while requiring orders-of-magnitude less computing resources and time. Open-access AIWP models thus hold promise as transformational tools for helping low- and middle-income populations make decisions in the face of high-impact weather shocks. Yet, current approaches to evaluating AIWP models focus mainly on aggregated meteorological metrics without considering local stakeholders' needs in decision-oriented, operational frameworks. Here, we introduce such a framework that connects meteorology, AI, and social sciences. As an example, we apply it to the 150-year-old problem of Indian monsoon forecasting, focusing on benefits to rain-fed agriculture, which is highly susceptible to climate change. AIWP models skillfully predict an agriculturally relevant onset index at regional scales weeks in advance when evaluated out-of-sample using deterministic and probabilistic metrics. This framework informed a government-led effort in 2025 to send 38 million Indian farmers AI-based monsoon onset forecasts, which captured an unusual weeks-long pause in monsoon progression. This decision-oriented benchmarking framework provides a key component of a blueprint for harnessing the power of AIWP models to help large vulnerable populations adapt to weather shocks in the face of climate variability and change.

cs.LG↗