How to Evaluate a Wind Site Without a Met Mast: 5 Methods Compared

Every wind farm starts with a question: is this site worth developing? Traditionally, answering that question meant installing a meteorological mast — a $50,000-$150,000 investment that takes 12-18 months to return bankable data. For the final investment decision, that level of rigor is non-negotiable. But for the 10-20 candidate sites you screen before choosing where to plant that mast? There are faster, cheaper ways to get a directionally accurate answer.

The wind industry has changed. Satellite-era reanalysis datasets now span four decades. Machine learning models trained on real production data can estimate capacity factors in minutes. Free government tools put global wind maps at your fingertips. Yet many developers still default to the same binary choice — expensive met mast or gut instinct — because they have not mapped out the full spectrum of preliminary wind resource assessment options available today.

This guide compares five methods for evaluating a wind site before committing to on-site measurement. Each has a role. None replaces a met mast for a final investment decision. But used at the right stage, any of them can save you months of time and tens of thousands of dollars in wasted campaigns.

Why Met Masts Are Not Always the Answer

Cost and timeline of met mast campaigns

A standard 80-meter met mast campaign costs between $50,000 and $150,000 when you factor in equipment, permitting, installation, data logger maintenance, and decommissioning. Taller masts — increasingly necessary as hub heights push past 120 meters — can run $200,000 or more. Remote or offshore sites add further cost through logistics and marine-grade instrumentation.

Timeline is the larger constraint. IEC 61400-12 standards require a minimum of 12 months of on-site measurement, and most lenders prefer 24 months. Add three to six months for permitting and installation, and you are looking at 18-30 months from decision to bankable dataset. During that window, land options can expire, grid interconnection queues can shift, and competing developers can lock up adjacent parcels.

The economics are clear: if you install met masts at 15 candidate sites to identify your best three, you have spent $750,000 to $2.25 million before a single turbine is ordered. Nobody does that. Instead, developers screen sites using desktop methods and reserve met mast budgets for sites that have already passed preliminary thresholds.

When you actually need a met mast vs. when you do not

A met mast (or LiDAR campaign) is essential when you need bankable energy estimates for financing — typically at the late feasibility or financial close stage. Lenders and equity investors require IEC-compliant measurement data, and no model-based assessment can substitute for it.

You do not need a met mast when you are:

• Scouting land — identifying regions or parcels with favorable wind characteristics
• Screening a portfolio — ranking 10-50 candidate sites to narrow down to 3-5
• Running preliminary economics — estimating whether a site's capacity factor can support a viable project at current PPA prices
• Responding to RFPs or tenders — providing indicative energy estimates within days, not months

For these early-stage activities, a wind resource assessment tool that delivers reasonable accuracy in hours or minutes is not just adequate — it is strategically superior because it compresses decision timelines.

Method 1 — Global Wind Atlas (Free)

The Global Wind Atlas (GWA) is a free, web-based platform developed by the Technical University of Denmark (DTU) in partnership with the World Bank. It provides modeled mean wind speed, power density, and wind rose data at a 250-meter resolution globally, using downscaled ERA5 reanalysis data combined with the WAsP microscale model.

Strengths:

• Completely free and accessible worldwide
• Provides a useful visual overview of wind resource patterns across regions
• 250-meter spatial resolution captures some terrain effects
• Includes power density and Weibull parameter estimates
• Useful for country-level or regional screening

Limitations:

• Reports only long-term mean values — no time series, no seasonal or diurnal profiles
• Cannot estimate capacity factor for a specific turbine model
• 250-meter resolution misses local effects in complex terrain (ridgelines, valleys, coastal acceleration zones)
• No production estimates, no AEP (annual energy production), no financial outputs
• Accuracy degrades significantly in areas with sparse surface station data (parts of Africa, Central Asia, and Southeast Asia)

The Global Wind Atlas is an excellent starting point for identifying promising regions, but it cannot tell you whether a specific site is commercially viable. Think of it as the wind equivalent of a topographic map — directionally useful, but not a survey.

Method 2 — NREL Wind Prospector (Free, US Only)

NREL's Wind Prospector is a GIS-based mapping tool that overlays modeled wind resource data across the continental United States, Hawaii, and parts of US territories. It draws on NREL's Wind Integration National Dataset (WIND Toolkit) and the AWS Truepower dataset, providing estimated wind speeds at multiple hub heights (typically 80m, 100m, and 140m).

Strengths:

• Free and well-maintained by a trusted government research lab
• Overlays wind data with transmission lines, substations, land ownership, and environmental constraints
• 2-km spatial resolution based on validated mesoscale modeling
• Includes capacity factor estimates at the county level
• Integrates with NREL's broader suite of tools (SAM, REopt, ATB)

Limitations:

• US only — not useful for international development portfolios
• 2-km resolution is too coarse for site-specific micrositing
• Data vintage varies; some layers have not been updated since 2012-2014
• County-level capacity factor averages mask significant intra-county variability
• No customization for specific turbine models or hub heights beyond preset options

For US-focused developers, Wind Prospector is a strong complement to GWA. It adds infrastructure context — proximity to transmission, land use classifications — that pure wind resource tools lack. But it remains a screening tool, not a site assessment tool.

Method 3 — ERA5 and MERRA-2 Reanalysis Data (Free, Requires Expertise)

ERA5 (produced by ECMWF) and MERRA-2 (produced by NASA) are global atmospheric reanalysis datasets that provide hourly wind speed, direction, temperature, pressure, and other variables on gridded meshes. ERA5 offers approximately 31-km resolution; MERRA-2 provides roughly 50-km resolution. Both span multiple decades (ERA5 from 1940 to near-present; MERRA-2 from 1980 onward), enabling robust long-term climatological analysis.

Strengths:

• Free to access through ECMWF's Climate Data Store (ERA5) or NASA's GES DISC (MERRA-2)
• Hourly time series enable diurnal and seasonal analysis, not just annual means
• Multi-decade records support long-term wind variability and trend analysis
• Global coverage with consistent methodology
• Widely accepted in the wind industry as the basis for measure-correlate-predict (MCP) workflows

Limitations:

• 31-50 km grid spacing is far too coarse for site-level assessment — a single grid cell may cover an area larger than most wind farm lease boundaries
• Wind speeds at hub height must be extrapolated from standard pressure levels, introducing vertical profile errors
• Reanalysis models smooth out terrain effects, systematically underestimating wind speeds on ridgelines and overestimating them in sheltered valleys
• Requires significant technical expertise in atmospheric science, Python/R scripting, and data processing
• Raw reanalysis data cannot produce a capacity factor estimate without additional turbine modeling

ERA5 and MERRA-2 are invaluable inputs to a wind site assessment workflow, but they are not assessments in themselves. Most professional wind resource assessment tools — including consultant desktop studies and machine learning platforms — use reanalysis data as foundational inputs and then add correction layers, downscaling, and turbine-specific modeling on top.

Method 4 — Consultant Desktop Study ($5,000-$15,000)

A consultant desktop study is a professional-grade preliminary wind resource assessment performed by a specialized consultancy (e.g., DNV, UL Renewables, Vaisala, or regional firms). The consultant typically combines reanalysis data, mesoscale modeling (WRF or ICON), and sometimes proprietary measurement databases to produce an estimated wind regime and indicative AEP for a specified site.

Strengths:

• Produced by experienced resource assessment engineers who apply professional judgment
• Typically includes mesoscale modeling at 1-3 km resolution, sometimes with simplified microscale adjustments
• Deliverable is a formal report suitable for internal investment committee review
• Can incorporate site-specific factors: terrain, roughness, wake effects from neighboring farms
• Consultants can adjust scope to match project-stage requirements

Limitations:

• Costs $5,000-$15,000 per site, making portfolio-wide screening expensive ($50K-$150K for 10 sites)
• Turnaround time ranges from 2-6 weeks depending on consultant backlog and scope
• Quality varies significantly between firms and even between analysts within the same firm
• Mesoscale models carry inherent uncertainty — typical bias ranges of 5-10% in complex terrain
• Each study is bespoke, making apples-to-apples comparison across sites difficult unless the same consultant assesses all of them

Desktop studies remain the industry standard for pre-feasibility assessment. They offer the best combination of rigor and speed short of actual measurement. But their cost and turnaround time make them impractical for early-stage land scouting or rapid portfolio screening where you need answers on 20 sites this week, not three sites this quarter.

Method 5 — WindAI Automated Assessment ($49.99, 2-5 Minutes)

WindAI is a machine learning-based wind resource assessment tool that generates site-specific capacity factor predictions from coordinates alone. The platform's model is trained on over 10 million hourly observations from 300+ operational wind farms across 8 countries, using 400+ engineered features derived from ERA5 reanalysis data, terrain analysis, and atmospheric boundary layer physics.

When you enter a site's latitude and longitude, WindAI extracts the same 400+ features for that location, runs them through its trained model, and returns a predicted net capacity factor — along with uncertainty bounds — in 2-5 minutes.

Strengths:

• Extremely fast: results in 2-5 minutes versus weeks (consultants) or months (met masts)
• Low cost: $49.99 per assessment, with the first assessment free
• Validated accuracy: RMSE of 0.147, R-squared of 0.777, and annual prediction errors of 2.1%-7.8% against operational wind farm data
• Consistent methodology across all sites, enabling true apples-to-apples portfolio comparison
• Incorporates terrain complexity, surface roughness, wind shear, atmospheric stability, and seasonal variability — factors that simpler tools ignore
• No software installation or GIS expertise required

Limitations:

• Currently trained on data from 8 countries — accuracy may degrade in regions with no training data coverage
• Predicts capacity factor, not a full wind regime (no wind rose, no directional analysis, no turbulence intensity)
• Does not replace a bankable energy assessment for financing purposes
• Model performance depends on the quality of underlying ERA5 data, which can be weaker in data-sparse regions
• Does not account for site-specific constraints like land exclusions, noise setbacks, or grid curtailment

Compared to Methods 1-3, WindAI adds the critical step those tools cannot: translating raw wind data into a production estimate calibrated against real-world wind farm performance. Compared to Method 4, it delivers a comparable screening-level output at roughly 1/100th the cost and 1/100th the turnaround time. The trade-off is depth — a consultant desktop study will include narrative context, sensitivity analysis, and professional sign-off that an automated tool does not.

Side-by-Side Comparison

Which Method Should You Use?

The right method depends on your development stage and what decision it needs to support. Here is a practical framework:

Land scouting (identifying regions of interest)

Use the Global Wind Atlas or WindAI. GWA gives you the broad geographic picture for free. WindAI lets you quickly test specific coordinates within promising regions to see if capacity factors are in a commercially viable range. At this stage, speed and coverage matter more than precision.

Initial screening (narrowing 20 sites to 5)

Use WindAI. This is where its combination of speed, cost, and site-specific output creates the most value. Running 20 assessments costs $1,000 and takes an afternoon. Running 20 consultant studies costs $100,000-$300,000 and takes two months. The free tools cannot produce the site-specific capacity factor estimates you need to make meaningful comparisons.

Pre-feasibility (building the initial business case)

Use a consultant desktop study or WindAI, depending on the decision's stakes. If you are presenting to an internal investment committee for the first time, a consultant report carries institutional weight. If you are making a go/no-go call on whether to spend $20,000 on a land option, WindAI's output may be sufficient and will arrive 10x faster.

Feasibility (supporting land agreements and permitting)

Use a consultant desktop study. At this stage, you need a professional-grade report that accounts for turbine-specific wake losses, terrain-driven uncertainty, and scenario analysis. Stakeholders and partners expect a formal deliverable with professional sign-off.

Bankable assessment (financial close)

A met mast or LiDAR campaign is required. No desktop method — whether consultant-led or AI-powered — substitutes for IEC-compliant on-site measurement when lenders are writing checks. This is where the $50,000-$150,000 investment is justified, because the project it supports is worth $50-$500 million.

The key insight is that these methods are not competitors — they are complementary stages in a pipeline. The most efficient developers use cheap, fast methods early to ensure that expensive, slow methods are only deployed on sites that have already cleared preliminary thresholds.

The Bottom Line

Evaluating a wind site without a met mast is not only possible — it is standard practice at every stage except financial close. The question is not whether to use desktop methods, but which desktop method matches your current decision point.

Free tools like the Global Wind Atlas and NREL Wind Prospector are valuable for broad-stroke regional screening. ERA5 and MERRA-2 reanalysis datasets provide the temporal depth needed for long-term variability analysis, provided you have the technical skills to work with them. Consultant desktop studies remain the gold standard for formal pre-feasibility work, delivering professional-grade assessments with expert narrative and sign-off.

What has been missing until recently is a fast, affordable, accurate option for the critical middle ground — the stage where you have 10-20 specific coordinates and need site-level capacity factor estimates this week, not this quarter. That is the gap that machine learning-based tools like WindAI are designed to fill: trained on 10 million+ real-world observations from 300+ wind farms, delivering validated predictions (R-squared of 0.777, annual errors of 2.1%-7.8%) in minutes instead of weeks.

No single method is sufficient for the full development lifecycle. But by matching the right tool to the right stage, you can move faster, screen more sites, and allocate your met mast and consulting budgets where they will generate the highest return.

Your first assessment is free at windai.tech.