3. Renewable Resource Assessment and Forecasting

Introduction and Context

The technical and economic performance of a renewable-energy project depends first on the quality of its resource assessment. A photovoltaic plant must be designed from solar irradiation rather than from nominal sunshine hours; a wind farm must be evaluated from a wind-speed distribution and turbine-specific power curve rather than from an annual average wind speed; and a hydropower scheme must account for seasonal flow, flood risk, sediment, and environmental release requirements. Similar reasoning applies to biomass, geothermal, and marine energy systems.

Resource assessment is therefore more than measurement. It is the systematic process of describing the magnitude, variability, spatial distribution, accessibility, uncertainty, and long-term evolution of an energy resource. Forecasting extends this process into the future. It supports dispatch, storage scheduling, market participation, maintenance planning, grid balancing, and investment decisions. The time horizon is important: a forecast for the next fifteen minutes is produced and evaluated differently from a climate-adjusted estimate of resource availability over the next thirty years.

A central engineering difficulty is that renewable resources are variable, partly predictable, and strongly influenced by location. Measurements are limited in duration and may contain instrument, sampling, and representativeness errors. The engineer must consequently combine ground observations, physical models, statistical methods, geographic information systems, remote sensing, and uncertainty analysis. The objective is not to produce a single apparently precise number, but to provide decision-makers with a defensible range of possible outcomes and the probability attached to them.

Learning Objectives

After studying this chapter, the student should be able to:

1. Explain the physical characteristics and principal assessment variables of solar, wind, hydropower, biomass, geothermal, and marine resources.

2. Select appropriate measurement systems, sampling procedures, and quality-control methods.

3. Apply probability distributions and statistical indicators to characterize resource variability.

4. Explain the roles of spatial data, remote sensing, mesoscale modelling, and resource atlases.

5. Distinguish short-, medium-, and long-term renewable-resource forecasting methods.

6. Quantify important sources of uncertainty and communicate results for engineering decisions.

7. Evaluate professional, environmental, regulatory, and data-governance issues in resource assessment.

Resource Assessment as an Engineering Decision Process

A bankable assessment normally proceeds through several connected stages: definition of the decision and study boundary; preliminary screening; measurement design; data collection; data validation; long-term correction or reconstruction; spatial extrapolation; energy-yield modelling; uncertainty quantification; and independent review. The required accuracy depends on the decision. Early site screening may tolerate broad regional estimates, whereas turbine financing, power-purchase negotiation, or grid interconnection requires documented measurements and an auditable methodology.

The resource variable must also be distinguished from energy production. Irradiance, wind speed, river discharge, fuel moisture, ground temperature, and wave height are resource variables. Power and energy are outputs obtained after applying a conversion model, equipment characteristics, losses, availability, and operational constraints. Confusing resource potential with deliverable energy is a frequent source of overestimation.

A measurement campaign should be designed around representativeness. Sensors must capture the vertical, horizontal, seasonal, and diurnal characteristics relevant to the proposed technology. The campaign should specify sensor type, mounting height, calibration, sampling interval, data logger requirements, redundancy, inspection frequency, communications, metadata, and procedures for missing data. A technically sophisticated model cannot compensate fully for poorly documented or non-representative observations.

Measurement quality and uncertainty

Common data-quality problems include sensor drift, icing, shading, fouling, communication failures, timestamp errors, unit inconsistencies, changes in land cover, and local obstructions. Quality control typically includes range tests, persistence tests, step-change tests, comparison with redundant instruments, physical consistency checks, and comparison with nearby stations or reanalysis products. Every correction should be logged rather than silently replacing the original value.

Useful statistical descriptors include the mean, median, standard deviation, coefficient of variation, quantiles, maximum, minimum, seasonal mean, and autocorrelation. For a variable x measured n times, the sample mean is

x̄ = (1/n) Σ xᵢ,

and the sample standard deviation describes dispersion around that mean. These statistics are necessary but not sufficient. Two sites can have the same annual mean wind speed or solar irradiation and very different hourly patterns, extreme events, and energy yields.

Uncertainty is often separated into measurement uncertainty, sampling uncertainty, model uncertainty, spatial uncertainty, and future-climate or long-term variability. A useful engineering presentation may report P50, P75, and P90 energy estimates. P50 is the median estimate; P90 is a conservative value that is expected to be exceeded with a specified probability under the adopted convention. Because conventions differ between industries, the probability definition must always be stated explicitly.

Solar Resource Characterization

Solar assessment begins with the distinction between irradiance and irradiation. Irradiance is instantaneous power per unit area, measured in W/m², whereas irradiation is energy per unit area over a period, commonly expressed in kWh/m²/day or kWh/m²/year. Global horizontal irradiance (GHI) consists of direct normal irradiance projected onto a horizontal surface and diffuse horizontal irradiance. Direct normal irradiance (DNI) is particularly important for concentrating solar power, while photovoltaic modules respond to the plane-of-array irradiance determined by orientation, tilt, tracking, ground reflection, and shading.

Pyranometers measure hemispherical short-wave irradiance on a plane, and pyrheliometers measure DNI through a narrow field of view, generally with solar tracking. Reference cells can provide a response closer to photovoltaic modules but are spectrally and technologically specific. Instruments require appropriate leveling, cleaning, ventilation where applicable, calibration, and protection from shading. Spectral effects, angle-of-incidence losses, soiling, temperature, and module degradation must be considered when converting irradiation to electricity.

Solar resource models frequently use satellite imagery, cloud properties, aerosol information, atmospheric transmittance models, and ground observations. Satellite data offer broad spatial coverage but may contain biases from cloud classification, aerosol assumptions, surface reflectance, and temporal sampling. Ground measurements provide local truth but are point observations. Combining both through bias correction or data assimilation is usually more reliable than using either source alone.

Solar forecasting is strongly linked to cloud movement. Very-short-term methods use sky imagers, irradiance persistence, and optical-flow tracking of clouds. Numerical weather prediction becomes increasingly useful from several hours to multiple days. For longer horizons, climatology, statistical weather generators, and climate projections describe expected distributions rather than exact hourly trajectories. Forecast error is commonly assessed using mean absolute error, root mean square error, mean bias error, and skill relative to a persistence or climatological benchmark.

Wind Resource Assessment

Wind energy assessment requires more than an average wind speed. Measurements should describe speed and direction at relevant heights, turbulence intensity, wind shear, air density, extreme gusts, and terrain-induced flow. Cup and sonic anemometers mounted on meteorological masts remain important reference instruments. LiDAR and SoDAR provide remote vertical profiles and can reduce the need for tall masts, although availability, atmospheric conditions, retrieval algorithms, and correlation with long-term measurements must be examined.

A wind-speed frequency distribution is often represented by the Weibull probability density function:

f(v) = (k/c)(v/c)^(k−1) exp[−(v/c)^k],

where v is wind speed, k is the shape parameter, and c is the scale parameter. The shape parameter indicates how concentrated the wind-speed observations are; the scale parameter is related to the characteristic speed. The Weibull model is convenient, but it may be inadequate where multiple wind regimes, complex terrain, seasonal monsoons, or strong directional effects exist. Measured distributions, mixture models, or sector-specific analysis may then be preferable.

The power available in moving air is proportional to air density and the cube of wind speed. Turbine output is determined by the power curve, which includes cut-in, rated, and cut-out speeds. Consequently, a small bias in wind speed can produce a disproportionately large error in predicted energy, particularly below rated speed. Air-density correction requires temperature, pressure, and sometimes humidity data. Terrain roughness, obstacles, wake losses, electrical losses, availability, curtailment, and degradation must be incorporated in the energy model.

Long-term correction commonly uses measure-correlate-predict (MCP) methods. A short on-site measurement record is correlated with a longer reference series from another mast, reanalysis dataset, or validated mesoscale model. Regression, matrix, Weibull, and machine-learning approaches may be used, but the correlation must be tested by season and wind sector. Poor correlation, structural changes at the reference site, or extrapolation beyond the observed range can create substantial uncertainty.

Wind forecasting uses persistence for the shortest horizon, statistical time-series methods for local correction, and numerical weather prediction for several hours to days. Ensemble weather forecasts provide a distribution of possible wind outcomes rather than one deterministic trajectory. Forecasts are valuable not only for energy trading but also for reserve procurement, storage control, curtailment planning, and maintenance scheduling.

Water, Hydropower, and Marine Resources

Hydropower assessment is based on the relationship between discharge and hydraulic head. The ideal hydraulic power is

P = ρgQHη,

where ρ is water density, g is gravitational acceleration, Q is discharge, H is net head, and η is overall efficiency. The usable energy depends on the flow-duration curve, which shows the percentage of time a particular discharge is equalled or exceeded. A plant designed solely for high flow may have low utilization during dry periods, while a plant sized for low flow may spill substantial water during floods.

Assessment requires stream-gauge records, rainfall data, catchment characteristics, topography, sediment information, and environmental-flow requirements. Short records may be extended using rainfall-runoff models or correlated nearby basins, but non-stationarity from land-use change, abstractions, regulation, and climate change must be considered. Flood frequency analysis is essential for hydraulic safety, while drought analysis informs firm energy and financial risk.

Marine resources include tidal range, tidal-stream currents, waves, and ocean thermal gradients. Tidal range can be assessed from harmonic analysis of water levels. Tidal-stream assessment requires current velocity profiles, bathymetry, turbulence, seabed conditions, and navigation constraints. Wave energy characterization uses significant wave height, peak or energy period, direction, joint probability distributions, and extreme-wave statistics. Buoys, acoustic Doppler instruments, radar, satellites, and hindcast models are complementary sources. Devices must be evaluated against survivability conditions as well as average energy conditions.

Unlike wind and solar, tidal currents are highly predictable through astronomical forcing, although turbulence, bathymetry, storms, and interactions with infrastructure introduce uncertainty. Wave forecasting uses spectral numerical models and data assimilation. Operational forecasts support deployment, access, and survivability decisions, while long-term hindcasts support site selection and structural design.

Biomass and Geothermal Resource Assessment

Biomass assessment concerns both quantity and quality. Relevant feedstocks include agricultural residues, forestry residues, energy crops, animal waste, municipal organic waste, and industrial by-products. A resource inventory must account for production rate, competing uses, collection radius, moisture content, ash, contamination, seasonal availability, storage losses, transport energy, and sustainability constraints. The theoretical biomass resource is therefore much larger than the technically and economically accessible resource.

Sampling should follow a defined protocol because moisture and composition vary spatially and temporally. Laboratory or field measurements may include proximate analysis, ultimate analysis, heating value, ash fusion behaviour, bulk density, and biochemical composition. GIS-based supply curves can link feedstock quantity and location with road networks, plant size, transport cost, and competing demand. Forecasts are influenced by agricultural production, land-use policy, climate, commodity prices, and waste-management practices.

Geothermal assessment begins with geological, geochemical, geophysical, and temperature-gradient evidence. Important variables include reservoir temperature, depth, permeability, fluid chemistry, pressure, recharge, and sustainable production rate. Exploration uncertainty is high because underground conditions are incompletely observed. Conceptual reservoir models are refined through well tests, tracer studies, pressure monitoring, and history matching. Probabilistic estimates of recoverable heat are more defensible than a single deterministic reserve figure.

Geothermal forecasting is less concerned with hourly weather variability than solar or wind forecasting. It focuses on reservoir response, well decline, reinjection, induced seismicity, and long-term thermal sustainability. Numerical reservoir simulation can test alternative production and reinjection strategies, but model results depend strongly on geological assumptions and calibration data.

Spatial Data, Remote Sensing, and Resource Mapping

Renewable resources are spatially heterogeneous. Geographic information systems allow the integration of satellite products, digital elevation models, land cover, protected areas, transmission lines, roads, substations, bathymetry, settlements, and hazard layers. A resource map becomes useful for planning only when technical, environmental, social, and grid constraints are overlaid.

Remote sensing includes satellite radiometers, synthetic aperture radar, LiDAR, thermal sensors, scatterometers, sky cameras, and airborne surveys. Its advantages are broad coverage, repeatability, and access to difficult terrain or offshore locations. Its limitations include retrieval bias, cloud obstruction, coarse resolution, temporal gaps, algorithm dependence, and the need for validation. Downscaling a regional product to a project site should not be treated as creating new information; it is a model-based estimate with associated uncertainty.

Mesoscale models simulate atmospheric or hydrological conditions over regional domains, while microscale models resolve local terrain and surface effects. A robust workflow often combines reanalysis or mesoscale fields with site measurements through bias correction, dynamical downscaling, or statistical post-processing. Cross-validation and independent holdout periods are necessary to prevent overfitting.

Forecasting Across Time Horizons

Forecasting methods should match the physical predictability and operational decision. The main horizons are:

• Very short term: seconds to approximately one hour, important for inverter control, ramp management, and immediate grid balancing. Persistence, sensor networks, sky imagery, and local state estimation are common.
• Short term: several hours to a few days, important for dispatch, reserve scheduling, and market bidding. Numerical weather prediction, statistical correction, ensemble methods, and hybrid machine-learning models are used.
• Medium term: days to weeks or months, useful for maintenance, fuel procurement, water management, and seasonal operation. Weather regimes, climate indices, hydrological forecasts, and probabilistic models are relevant.
• Long term: years to decades, required for project finance, resource planning, and climate resilience. Climatology, reconstructed historical series, scenario analysis, and climate projections are used.

Persistence is a strong benchmark for short horizons because renewable output often changes gradually. Autoregressive and state-space models capture temporal dependence, while regression models incorporate explanatory variables such as cloud cover, pressure gradients, temperature, and numerical weather predictions. Machine-learning models, including gradient-boosting methods and recurrent neural networks, can improve nonlinear prediction when sufficient high-quality training data are available. They may nevertheless fail under unprecedented weather, changing sensor behaviour, or a shift in the data distribution.

Probabilistic forecasting produces prediction intervals, quantiles, or ensembles. These are often more useful than a point forecast because operators must decide how much reserve or storage capacity to hold. Forecast quality should therefore be evaluated for both sharpness and calibration: a narrow interval is valuable only if the stated coverage probability is achieved. Economic metrics, such as imbalance cost or avoided reserve cost, complement statistical metrics.

Applied Case: Illustrative Wind-Solar Hybrid Assessment

Background

Consider a proposed 100 MW hybrid project in a semi-arid region with a 60 MW photovoltaic plant and a 40 MW wind plant. The developer has one year of solar measurements, eighteen months of mast data, satellite irradiation data, a nearby long-term wind reference series, and preliminary grid-connection information. The project is intended to supply a firm evening-oriented profile using a battery system.

Situation and problem

The annual mean solar resource appears favourable, but the site experiences monsoon cloud bands during a period when wind speeds are also uncertain. The developer must estimate annual energy, hourly coincidence between solar and wind generation, battery size, and the probability of failing a contracted delivery profile. A simple addition of independently modelled P50 outputs would ignore correlation and could understate risk.

Analysis and approach

First, the measurement records are quality-controlled. Periods affected by sensor fouling and mast communication failure are flagged, not automatically filled. The solar series is compared with satellite estimates and corrected using a seasonally stratified bias model. The wind series is correlated with the long-term reference data by wind direction and month. Wind shear is estimated using simultaneous measurements at multiple heights and checked against a terrain-flow model.

Next, hourly solar and wind series are reconstructed for a long historical period. The project team generates correlated weather scenarios rather than sampling solar and wind independently. PV output is calculated using plane-of-array irradiation, module temperature, inverter clipping, soiling, and availability. Wind output uses air-density-adjusted turbine curves, wake losses, electrical losses, and curtailment assumptions. Battery simulations then test alternative capacities under degradation and round-trip-efficiency assumptions.

Uncertainty is represented through distributions for long-term bias, wind-speed adjustment, availability, degradation, wake loss, and forecast error. The team reports P50 and conservative exceedance cases, while also presenting seasonal energy and ramp statistics. A sensitivity analysis shows that uncertainty in long-term wind conditions and wake losses has a greater effect on firm evening delivery than uncertainty in PV module efficiency.

Outcome

The assessment indicates that the hybrid plant has greater profile value than either resource alone, but the battery requirement depends strongly on the joint distribution of cloud and wind events. The project is not approved solely on annual P50 energy. Instead, the grid contract is revised to include a realistic delivery tolerance, a reserve requirement, and curtailment rules. Additional wind measurement and one more year of solar validation are recommended before final investment decision.

Learning points

The case demonstrates that resource assessment must connect physical measurements to the actual commercial decision. Correlation between resources, not only their individual averages, determines hybrid-system value. Transparent assumptions, probabilistic results, and sensitivity analysis are more useful than a single optimistic yield estimate.

Challenges, Governance, and Professional Practice

The major technical challenge is limited representativeness. A short measurement record may coincide with an unusual year, while a long reference record may not represent the project site. Climate variability and change challenge the assumption of stationarity. Data gaps, model bias, and uncertain future land use add further complexity.

Professional practice requires traceable metadata, version-controlled datasets, documented algorithms, calibration certificates, and reproducible calculations. Data ownership and cybersecurity matter when sensors are connected through remote networks. Forecasts should not be presented with unjustified precision, particularly when they influence public infrastructure, land acquisition, or financing.

Environmental and social governance is integral to assessment. Hydropower studies must address aquatic connectivity, sediment transport, fisheries, and downstream flows. Biomass studies must avoid indirect land-use change, food-versus-fuel conflicts, and unsustainable removal of residues. Geothermal projects require attention to fluid disposal and induced seismicity. Wind, solar, and marine projects require assessment of biodiversity, visual effects, navigation, noise, cultural heritage, and community concerns. Regulatory requirements vary by jurisdiction, but compliance should be incorporated at the screening stage rather than after technical optimization.

Current Developments and Best Practices

Digital resource assessment increasingly uses cloud-based data platforms, automated quality control, open geospatial layers, digital twins, and probabilistic simulation. Artificial intelligence is being applied to satellite-image interpretation, gap filling, forecast post-processing, anomaly detection, and short-term prediction. Its use should remain physically informed: an accurate model in historical data can fail when the weather regime, equipment, or sensor network changes.

Best practice can be summarized as follows:

1. Define the decision, time horizon, and required confidence before collecting data.

2. Combine independent data sources and validate remote or modelled products against ground observations.

3. Preserve raw data and document every correction, filtering rule, and imputation method.

4. Model temporal and spatial correlation, especially for hybrid plants and regional portfolios.

5. Report uncertainty, sensitivity, and extreme conditions rather than only annual averages.

6. Use benchmarks and out-of-sample testing for forecasting models.

7. Include environmental, social, regulatory, and grid constraints in site assessment.

8. Reassess the resource as new measurements, operational data, and climate information become available.

Chapter Summary

Renewable-resource assessment converts measurements and models into an evidence base for engineering and investment decisions. Solar assessment distinguishes GHI, DNI, and plane-of-array irradiation; wind assessment uses vertical profiles, turbulence, air density, power curves, and long-term correlation; hydropower relies on head, discharge, flow-duration curves, floods, and droughts; marine assessment addresses tides, currents, waves, and survivability; biomass requires a sustainable, spatially explicit supply analysis; and geothermal assessment depends on uncertain subsurface conditions and reservoir behaviour.

Probability distributions, time-series statistics, GIS, remote sensing, reanalysis, mesoscale models, and uncertainty analysis are complementary tools. Forecasting must be matched to its horizon, from persistence and sky imaging to numerical weather prediction, ensemble forecasting, seasonal methods, and climate scenarios. A professional assessment is transparent, validated, probabilistic, and connected to the actual operational or financial decision. It also recognizes environmental obligations, data governance, model limitations, and the risks of false precision.

Key Terms

Irradiance: Instantaneous solar power received per unit area.

Irradiation: Solar energy received per unit area over a specified period.

Global horizontal irradiance: Total direct and diffuse solar irradiance incident on a horizontal plane.

Direct normal irradiance: Direct solar irradiance received on a plane normal to the sun’s rays.

Wind shear: Variation of wind speed or direction with height above the ground or surface.

Turbulence intensity: Ratio of wind-speed standard deviation to mean wind speed.

Weibull distribution: A probability distribution frequently used to represent wind-speed frequency.

Measure-correlate-predict: A method for extending a short site measurement record using a longer correlated reference record.

Flow-duration curve: A curve showing the percentage of time that a river discharge is equalled or exceeded.

Hydraulic head: The effective elevation or pressure difference available to drive hydropower production.

Hindcast: A model-generated reconstruction of historical environmental conditions.

Reanalysis: A physically consistent historical dataset produced by assimilating observations into a numerical model.

Mesoscale model: A numerical model describing atmospheric or environmental conditions over regional spatial scales.

Remote sensing: Measurement of environmental variables without direct physical contact with the observed location.

P50 estimate: A median or central estimate under a stated probability convention.

Prediction interval: A range intended to contain a future observation with a specified probability.

Bias correction: A statistical or physical adjustment intended to reduce systematic error in a model or reference dataset.

Representativeness error: Error arising when a measurement does not adequately describe the location, period, or scale of interest.

Resource atlas: A spatial database or map showing the distribution and characteristics of an energy resource.

Firm energy: Energy that can be delivered with a specified level of reliability under resource and operating constraints.

Review Questions

Short Answer Questions

1. Distinguish irradiance from irradiation and explain why both are used in solar assessment.

2. What information does a flow-duration curve provide for a hydropower project?

3. State two reasons why an annual average wind speed is insufficient for turbine-yield estimation.

4. What is the purpose of quality control in a renewable-resource measurement campaign?

5. Define prediction interval and explain its value to a grid operator.

Long Answer Questions

1. Describe a complete workflow for assessing a wind-energy site from measurement design to probabilistic annual energy estimation.

2. Explain how satellite data, ground measurements, reanalysis, and mesoscale models can be combined in solar-resource assessment.

3. Compare the principal resource variables, measurement systems, and major uncertainties for solar, wind, and hydropower projects.

4. Discuss short-, medium-, and long-term renewable-energy forecasting methods and the decisions supported by each horizon.

5. Explain why biomass and geothermal assessments require sustainability and subsurface or supply-chain analysis in addition to resource magnitude.

Analytical / Application Questions

1. A site has the same annual mean wind speed as a reference site but a much higher frequency of calm periods and extreme gusts. Explain how this could affect turbine selection, energy yield, and structural design.

2. A photovoltaic project has excellent P50 annual irradiation but large uncertainty in cloud-related evening ramps. Propose a measurement, forecasting, and battery-sizing strategy.

3. A hydropower developer has only five years of discharge data in a basin undergoing land-use change. Identify the principal uncertainties and recommend a defensible assessment approach.

22. Multiple Choice Questions with Answers

Question 1

Which quantity is most directly required to evaluate the hydraulic power potential of a river site?

A. Solar azimuth and albedo

B. Discharge and net hydraulic head

C. Wind direction and turbulence intensity

D. Biomass ash content and bulk density

Correct Answer: B. Discharge and net hydraulic head

Question 2

Why can a small wind-speed bias produce a relatively large error in predicted wind-turbine energy?

A. Wind speed affects power approximately through a cubic relationship below rated conditions

B. Wind speed determines only the turbine’s electrical frequency

C. Wind speed has no relationship with air density

D. Wind speed changes the solar altitude angle

Correct Answer: A. Wind speed affects power approximately through a cubic relationship below rated conditions

Question 3

Which method is specifically used to extend a short on-site wind record with a longer reference dataset?

A. Harmonic tidal analysis

B. Measure-correlate-predict

C. Proximate biomass analysis

D. Flow-duration transformation

Correct Answer: B. Measure-correlate-predict

Question 4

A solar engineer wants to assess a concentrating solar-power plant. Which resource variable is especially important?

A. Diffuse horizontal irradiance only

B. Direct normal irradiance

C. River sediment concentration

D. Biomass moisture content

Correct Answer: B. Direct normal irradiance

Question 5

What is the principal advantage of probabilistic forecasting over a single point forecast?

A. It eliminates all measurement errors

B. It provides possible outcomes and uncertainty for operational decisions

C. It guarantees that the highest forecast will occur

D. It removes the need for historical data

Correct Answer: B. It provides possible outcomes and uncertainty for operational decisions

Question 6

Which statement best describes a flow-duration curve?

A. It gives the direction of wind at each turbine height

B. It shows the percentage of time a specified river discharge is equalled or exceeded

C. It measures the spectral response of a photovoltaic module

D. It predicts the depth of a geothermal reservoir

Correct Answer: B. It shows the percentage of time a specified river discharge is equalled or exceeded

Question 7

A resource map overlays irradiation, transmission lines, protected areas, roads, and settlements. This illustrates the use of GIS for:

A. Only instrument calibration

B. Integrated technical and site-constraint screening

C. Replacing all field measurements

D. Determining turbine cut-in speed directly

Correct Answer: B. Integrated technical and site-constraint screening

Question 8

Which issue is particularly important when estimating accessible biomass potential?

A. The fact that all theoretical biomass is economically and sustainably collectable

B. Competing uses, moisture, transport distance, and storage losses

C. The astronomical predictability of tides

D. The optical tracking of clouds

Correct Answer: B. Competing uses, moisture, transport distance, and storage losses

Question 9

Why should a hybrid solar-wind assessment model correlation between the two resources?

A. Correlation determines the combined temporal profile and storage requirement

B. Correlation changes the rated capacity printed on turbine nameplates

C. Correlation removes the need for battery degradation modelling

D. Correlation makes environmental assessment unnecessary

Correct Answer: A. Correlation determines the combined temporal profile and storage requirement

Question 10

Which practice most improves the credibility of a renewable-resource assessment?

A. Reporting only the most favourable annual average

B. Hiding corrected data to avoid confusion

C. Documenting assumptions, validating datasets, and reporting uncertainty

D. Using a machine-learning model without out-of-sample testing

Correct Answer: C. Documenting assumptions, validating datasets, and reporting uncertainty