When Does Advection-Aware Graph Nowcasting Help? A Controlled Study of Distributed Solar Ramp Forecasting with a Self-Supervised Cloud-Motion Estimator

Phillip Jiang

cs.LGSep 8, 2026

When Does Advection-Aware Graph Nowcasting Help? A Controlled Study of Distributed Solar Ramp Forecasting with a Self-Supervised Cloud-Motion Estimator

Authors: Phillip Jiang

Organizations: Appsofa LLC

Abstract

Short-term forecasting of cloud-induced power ramps across a network of distributed photovoltaic (PV) or irradiance sensors is a recognised pain point for grid operators. A natural idea is to make the graph neural network (GNN) advection-aware: connect each site to the sites upwind of it, with edge time-lags set by the cloud-motion vector (CMV), so that a ramp is propagated forward before it physically arrives. Using a controlled synthetic testbed with a known wind field, we show that (i) with a realistic cross-correlation CMV estimate, an explicit advection graph does not beat a plain static or learned-adjacency spatiotemporal GNN; (ii) roughly half of the benefit available from a perfect CMV comes simply from providing an accurate motion vector as an input feature, not from graph structure; and (iii) advection helps only when the advective displacement over the forecast horizon, v*H, fits inside the sensor network. Motivated by (ii), we introduce a small self-supervised cloud-motion estimator -- a position-aware encoder trained only on a multi-lag optical-flow reconstruction objective with an annealed kernel -- that recovers the true wind vector to 2-4 degrees median angular error, 2-4x better than the classical cross-correlation method across every wind regime. Freezing this estimator and feeding its vector to the forecaster closes about 60% of the oracle-CMV RMSE gap at moderate wind (8-15% RMSE reduction over no advection), with no external wind data. We also report a negative result for a spatially-coherent probabilistic head. All claims are established on a single synthetic simulator; we discuss why real-network validation is the necessary next step and outline it.

Figures & tables

Figure 1: E1. Test RMSE vs. wind speed, estimated CMV, 2 seeds (

\pm

s.d.). The advection feature (top line) does not beat the heterogeneous GNN without it, nor the learned-adjacency baseline.

	wind speed (m/s)
model	4	10	16
per-site GRU	0.066	0.148	0.154
static-distance GCN	0.034	0.089	0.130
learned adjacency	0.043	0.093	0.120
hetero GNN (no adv.)	0.032	0.087	0.127
hetero GNN $+$ advection	0.033	0.095	0.133

Table 1: E1 test RMSE (clear-sky index), estimated CMV, 2 seeds. Lower is better; best per column in bold .

wind	CMV	no adv.	$+$ adv.	$\Delta$ RMSE
$10\text{\,}\mathrm{m}\mathrm{/}\mathrm{s}$	estimated	0.086	0.096	$+11\%$
$10\text{\,}\mathrm{m}\mathrm{/}\mathrm{s}$	oracle	0.077	0.066	$-14\%$
$16\text{\,}\mathrm{m}\mathrm{/}\mathrm{s}$	estimated	0.133	0.128	$-4\%$
$16\text{\,}\mathrm{m}\mathrm{/}\mathrm{s}$	oracle	0.124	0.103	$-17\%$

Table 2: E2. Advection feature value depends on CMV quality (test RMSE, 2 seeds).

Figure 2: Self-supervised cloud-motion estimation. Median angular error of

\hat{\mathbf{v}}

vs. the true wind, 3 seeds. The learned estimator beats the classical cross-correlation method in every wind regime.

regime	learned (recon)	learned ( $+$ light sup.)	classical
steady, 4–16 m/s	3.9–7.5	2.0–3.1	14–31
variable, 4–16 m/s	8.5–12.8	6.9–7.7	15–25

Table 3: CMV estimator median angular error (deg), 3 seeds.

Figure 3: Frozen two-stage forecaster , 3 seeds (

\pm

s.d.). Green labels: fraction of the no-advection

\to

oracle gap closed. In the shaded region (

v\cdot H

off-grid) even the oracle does not beat no advection.

wind (m/s)	no adv.	frozen CMV	oracle	CMV err. (deg)
8	0.078	0.069	0.064	2.8
12	0.114	0.098	0.085	3.9
16	0.139	0.155	0.147	2.9
20	0.172	0.188	0.186	2.4

Table 4: Frozen two-stage test RMSE (clear-sky index), 3 seeds.

Explore similar work

Oct 13, 2026

PING: physics informed graph neural network for advection-diffusion processes with applications to solar power forecasting

Advection-diffusion plays an essential role in transport phenomena such as atmospheric and cloud dynamics. Improved forecasting of cloud motion, in turn, enhances solar power production forecasts and facilitates the integration of additional photovoltaic (PV) sources into the power grid. However, both classical numerical models and emerging deep learning approaches face challenges when applied to spatially irregular sensor networks such as those formed by weather stations. In this work, we adopt a graph signal processing perspective and model advection-diffusion particles as signals on a graph to capture the dynamics of these processes. We introduce a Physics-Informed Graph Neural Network (PING) for forecasting future advection-diffusion particle concentrations on irregular grids. The model incorporates a discretized advection-diffusion equation as a soft constraint in the loss function to guide learning toward physically consistent solutions. This model is on par with state of the art models on synthetic and real transport phenomena datasets, including sea surface temperature and cloud index datasets. Furthermore, the PING model is successfully coupled with models for forecasting PV power production. This hybrid approach leverages data fusion from ground-based sensors at PV stations and satellite images with cloud concentration to improve the accuracy of intra-day PV power production.

Jelena Simeunović, Baptiste Schubnel, Pierre-Jean Alet +2

CSEM, Rue de Jaquet-Droz 1, 2002 Neuchâtel, Switzerland · Department of Electrical Engineering, École Polytechnique Fédérale de Lausanne (EPFL), Rte Cantonale, 1015 Lausanne, Switzerland.

Physics-Informed GNNs Spatiotemporal Forecasting Solar Power Forecasting

Ask PDF

Apr 10, 2026

On-Meter Graph Machine Learning: A Case Study of PV Power Forecasting for Grid Edge Intelligence

This paper presents a detailed study of how graph neural networks can be used on edge intelligent meters in a microgrid to forecast photovoltaic power generation. The problem background and the adopted technologies are introduced, including ONNX and ONNX Runtime. The hardware and software specifications of the smart meter are also briefly described. Then, the paper focuses on the training and deployment of two graph machine learning models, GCN and GraphSAGE, with particular emphasis on developing and deploying a customized ONNX operator for GCN. Finally, a case study is conducted using real datasets from a village microgrid. The performance of the two models is compared on both the PC and the smart meter, exhibiting successful deployments and executions on the smart meter.

Jian Huang, Zixiang Ming, Yongli Zhu +1

Sun Yat-sen University, Guangzhou, China

Edge ML Solar Power Forecasting Graph Neural Networks

Ask PDF

Mar 21, 2026

Cloud-Edge Collaborative Large Models for Robust Photovoltaic Power Forecasting

Photovoltaic (PV) power forecasting in edge-enabled grids requires balancing forecasting accuracy, robustness under weather-driven distribution shifts, and strict latency constraints. Existing models work well under normal conditions but often struggle with rare ramp events and unexpected weather changes. Relying solely on cloud-based large models often leads to significant communication delays, which can hinder timely and efficient forecasting in practical grid environments. To address these issues, we propose a condition-adaptive cloud-edge collaborative framework CAPE for PV forecasting. CAPE consists of three main modules: a site-specific expert model for routine predictions, a lightweight edge-side model for enhanced local inference, and a cloud-based large retrieval model that provides relevant historical cases when needed. These modules are coordinated by a screening module that evaluates uncertainty, out-of-distribution risk, weather mutations, and model disagreement. Furthermore, we employ a Lyapunov-guided routing strategy to dynamically determine when to escalate inference to more powerful models under long-term system constraints. The final forecast is produced through adaptive fusion of the selected model outputs. Experiments on two real-world PV datasets demonstrate that CAPE achieves superior performance in terms of forecasting accuracy, robustness, routing quality, and system efficiency.

Nan Qiao, Shuning Wang, Sijing Duan +5

School of Computer Science and Engineering, Central South University, Changsha, Hunan 410083, China · Department of Computer Science and Technology, BNRist, Tsinghua University, Beijing 100084, China · Beijing Smartchip Microelectronics Technology Co., Ltd, Beijing 102200, China

Solar Power Forecasting Distribution Shift Robustness Adaptive Model Routing Edge Computing Retrieval-Augmented Time Series Forecasting

Ask PDF