AORCviz User Guide

What is AORCviz?

AORCviz is an interactive visualization tool for exploring meteorological data across the continental United States. It integrates two primary gridded datasets, real-time USGS stream gauge observations, and GHCN-D weather station data — all in one browser-based portal. An optional AI Chat assistant can drive the portal for you, and a Google Earth Engine connection adds satellite imagery, elevation, and global environmental datasets.

Datasets

AORC v1.1 (Historical) — NOAA's Analysis of Record for Calibration. 1 km resolution, hourly, 1979–present. Ideal for historical event analysis and climatology.

HRRR (Forecast) — NOAA's High-Resolution Rapid Refresh model. 3 km resolution, updated every hour, Aug 2016–present. Extended runs (00z, 06z, 12z, 18z) produce 48-hour forecasts; all other hours produce 18-hour forecasts. Includes surface and upper-air (pressure level) variables.

Switch between datasets using the AORC / HRRR toggle at the top of the sidebar.

Quick Start — AORC

1
Set your bounding box — Choose a center location and box size (km), or use preset buttons. Click "⊕ Click Map to Set" to pick a center anywhere on the map.
2
Select variables — Add up to 3 layers (Precipitation, Temperature, Wind, etc.) with independent opacity controls.
3
Choose dates — Enter start/end dates, or click a notable event preset (e.g., Hurricane Harvey 2017). Switch to Multi-Year Seasonal mode for climatology.
4
Fetch Data — The aggregated map renders. Switch to Time Slider to animate hourly timesteps. Click any grid cell for a time series chart.

Quick Start — HRRR

1
Switch to HRRR — Click the "HRRR (Forecast)" button. The latest available initialization date and hour are auto-detected.
2
Add layers — Select up to 3 variables from surface or pressure levels (500mb, 700mb, 850mb, 250mb jet stream, etc.).
3
Fetch Data — The aggregated map shows the full forecast summary. Use the Time Slider to step through individual forecast hours. Extended runs (00/06/12/18z) have 48 hours; other runs have 18 hours.
4
Click any point — See the full forecast time series for that location.

Observation Layers

Toggle observation markers using the buttons at the bottom-left of the map:

Stream Gauges (blue dots) — USGS real-time streamflow gauges. Click any gauge to see a 7-day hydrograph with statistics (min, max, mean, std). Use the period buttons (1d, 3d, 7d, 14d, 30d) to change the time window. In AORC mode, gauges automatically use the selected date range.

Weather Stations (orange squares) — GHCN-D official stations (COOP + first-order). Click a station to see recent observations, statistics, monthly normals, and charts. Select which variables to display and adjust the day range with the slider. Data comes from NOAA's CDO API (near real-time).

Both observation layers include a ⬇ CSV download link in the popup for exporting data.

AORC Variables

Precipitation (mm) · Temperature (°C) · Specific Humidity (kg/kg) · Downward LW/SW Radiation (W/m²) · Surface Pressure (Pa) · Wind Speed & Direction (m/s)

HRRR Variables

Surface: Temperature, Dew Point, Humidity, Wind, Gust, Precipitation, Pressure, MSLP, Radiation, CAPE, Reflectivity, Visibility

Pressure Levels: Temperature, Dew Point, Geopotential Height, Wind, Vertical Velocity — available at 250, 300, 500, 700, 850, 925, 1000 mb

AI Chat (Beta)

Open the AI Chat from the tab bar (next to User Guide) to analyze data conversationally. The AI can drive the portal itself: set the bounding box, fetch AORC or HRRR data, find and query stream gauges and weather stations, open time-series popups on the map, build charts in the chat, and offer downloads — just describe what you want, e.g. "Show me the precipitation during the July 2025 Texas floods and compare it against nearby station observations."

1
Add your API key — Click ⚙ Settings inside the chat and choose OpenAI, Anthropic, Google Gemini, or an open-source model via any OpenAI-compatible endpoint (Groq, Together, Ollama, LM Studio…). Paste your API key (not needed for local servers). The key is stored only in your browser and sent directly to the LLM provider — never to our server.
2
Ask a question — The AI fetches and analyzes data step by step; the maps update live as it works. Charts it creates can be saved as PNG, and it can package results as CSV, JSON, NetCDF (AORC), or GeoTIFF (Earth Engine).
3
Move & resize — Drag the chat by its header to see the map behind it, and use ⛶ to cycle panel sizes.
Larger models (GPT-4o, Claude Sonnet/Opus, Gemini Pro) handle multi-step analyses much better than the fast/mini tiers.

Earth Engine Layers (Optional)

Connect your own Google Earth Engine account (⚙ Settings in the AI Chat: OAuth Client ID + EE-enabled GCP project) to unlock satellite and global datasets: Sentinel-2 and Landsat imagery, SRTM elevation, land cover, GPM precipitation, JRC surface water, and more.

The AI can overlay EE results directly on the map, play animated time-lapses (e.g., monthly imagery through a flood event), compute zonal statistics, and produce GeoTIFF downloads. Requires an Earth Engine-registered Google account; sign-in happens entirely in your browser.

Added layers appear in the EE Layers panel on the map — a GIS-style layer control where you can toggle visibility, adjust opacity, or remove each layer. Drag the panel by its header to reposition it; click the header to collapse the list. EE layers stay visible when you switch basemaps.

Try: "Add SRTM elevation to the map", "Show a Sentinel-2 time-lapse of June–August 2025", or "Overlay flood-prone areas from JRC surface water".

Tips

Toggle between Metric (°C, mm, m/s) and Imperial (°F, in, mph) units in the sidebar. All temperatures display in °C by default.

Times are shown in local time based on the bounding box center — no manual timezone conversion needed.

The Time Slider colorbar stays fixed across all frames for consistent visual comparison.

Larger bounding boxes (500+ km) take longer to load. The AORC Time Slider supports up to 62 days; HRRR shows 48 forecast hours for extended runs (00/06/12/18z) and 18 hours for all other runs.

Overlay multiple variables with different opacities to compare patterns — e.g., 850mb temperature + surface wind + precipitation.

Export AORC results as PNG or NetCDF from the sidebar. Observation popups include CSV download links.

Data Sources

AORC v1.1: s3://noaa-nws-aorc-v1-1-1km (NOAA AWS Open Data)

HRRR: s3://hrrrzarr (University of Utah / NOAA AWS Open Data)

Stream Gauges: USGS National Water Information System (waterservices.usgs.gov)

Weather Stations: NOAA GHCN-D via CDO Web Services + S3 archive

Earth Engine (optional): Google Earth Engine data catalog, via your own EE account

NASA Earthdata (optional): NASA satellite products via CMR search + cloud subsetting, with your own Earthdata token

AI Chat (optional): OpenAI, Anthropic, Google Gemini, or any OpenAI-compatible endpoint (Groq, Together, Ollama, …), via your own API key — keys and requests never pass through our server

AORC
AORCviz
NOAA · AORC v1.1 · 1 km · Hourly · 1979–present
TExUS Lab, Jackson School of Geosciences, UT Austin
API: checking…
s3://noaa-nws-aorc-v1-1-1km
250 km box
Created by Naveen Sudharsan (naveens@utexas.edu)
Aggregated Map
Time Slider
AI Chat
User Guide
Stream Gauges
Weather Stations
Lat: —   Lon: —
Loading…
Lat: —   Lon: —
Lat: —   Lon: —
Loading…
AI Data Analysis
Configure your API key in ⚙ Settings, then ask anything — e.g. "Show me the precipitation during the July 2025 Texas floods". The AI can set the map region and fetch the data itself.