Read your site: sun, frost dates and climate
How to measure direct sun on your spot, read your frost dates as probabilities, and see what a hardiness zone does and does not say about vegetables.
- Start here: the essentials
- 1. Understand: light is the garden's fuel
- Measuring direct sun: the sun map
- Microclimates: cold air, heat and wind
- What a hardiness zone measures, and what it does not
- Frost dates are probabilities
- Where frost is rare, heat sets the season
- 2. See it
- 3. Practice
- 4. Your garden
- 5. Check your work
- 6. Knowledge check and scenario
Start here Essentials
This module asks for two measurements: the hours of direct sun where the garden will sit, and your frost dates with their source. Everything marked Go deeper is optional on a first pass.
- Make one sun map. On a clear day, look at each corner and the center of the spot every hour and write S (direct sun), D (dappled) or N (shade). Count the S marks. A spring map is a first measurement; summer leaves and sun can change it. The sun map.
- Record the lowest count as the spot's sun hours, with the date. Give the sunniest end to fruiting crops. Reading the result.
- Look up your frost dates. For the nearest station, find the 50 percent and 10 percent dates for the last spring and first fall 32 °F freeze. Write the station, threshold and data period. Where to look up your dates. If your area rarely frosts, read the frost-free section instead.
- Choose and enter your dates. The 50 percent date is the median. The 10 percent date is the lower-risk choice. Write the reason and enter both dates in setup.
Minimum time: about 15 minutes across one clear day, plus 30 minutes for the frost lookup. Done when: the spot's sun hours and your frost dates, with their source, are written down and entered in setup.
1. Light is the garden's fuel Understand
Two facts about a site decide more of a first garden's success than any other: how many hours of direct sun the plants get, and how long the frost-free season lasts. Neither can be changed much by effort. You can water, feed and weed your way around many problems, but you cannot add hours of sun to a spot beside a north wall, and you cannot move your frost dates. So the sensible order is to measure both before you build, plant or buy, and let the measurements choose the spot and the crops.
Physical principle Plants build everything they are made of from carbon dioxide and water, using the energy in light. More light, up to a point, means more sugar to build with. A leafy crop such as lettuce is harvested as leaves, which the plant makes early and cheaply. A fruiting crop such as a tomato has to grow leaves, stems and roots first and then fill flowers and fruit, which are expensive in sugar. That is why extension services ask more sun for fruiting crops than for leafy and root crops, and why a fruiting crop in too little light often grows plenty of leaves and little fruit.
Direct sun means the sun itself is visible from the spot, so the plants cast a sharp shadow. Bright shade (open sky but no sun, such as the north side of a house on a clear day) and dappled shade under a thin canopy are both less. When extension services give hours of sun, they mean direct sun.
| Crop type | Hours of direct sun | Source |
|---|---|---|
| Fruiting crops (tomato, pepper, eggplant, squash, cucumber, bean) | At least 6 to 8 a day | Oregon State University Extension, EM 9032 |
| Salad and leafy greens, radishes, turnips, beets | 4 to 6 a day, or constant dappled shade | Oregon State University Extension, EM 9032 |
| Collards, kale, mustard, turnip greens, pac choi | Tolerate partial shade, 4 to 6 hours of direct light | University of Maryland Extension |
| A whole vegetable garden in South Carolina | At least 6; leafy vegetables tolerate partial shade | Clemson Cooperative Extension, HGIC 1256 |
Convention These are conventions from extension services, not thresholds where a plant switches on. Below them, a crop may grow but produce less. Oregon State's 6 to 8 hours for fruiting crops and its 4 to 6 hours for greens and roots are the figures this course uses. Test it If your best spot measures five hours, a tomato there is a question your own harvest log answers, not a rule that forbids it. The useful decision is placement: give the sunniest beds to the fruiting crops, and put lettuce, spinach, kale and radish where the light is shorter.
Placement within the garden matters too. Convention Clemson Extension advises putting tall and trellised crops on the north side of the garden so they do not shade the shorter ones; in the southern hemisphere the same logic puts them on the south side. Clemson also advises keeping the garden away from trees and shrubs, which compete for water and nutrients as well as light.
Measuring direct sun: the sun map
People are poor judges of sun hours. Memory keeps the bright midday and forgets that the sun left the bed at 2 pm behind the garage. A sun map is a sketch of the candidate spot with the hours of direct sun recorded at several points, made by checking the spot at fixed times through one day. It replaces a guess with a count.
How to make one
Sketch the spot from above, roughly to scale: the bed outlines or the area you might dig, the house, fences, trees and anything tall nearby. Mark three to five check points: each corner of a planned bed and its center is a good set. Then, on a clear day, look at each point once an hour from early morning to evening. At each check, write one letter beside the point: S if the sun is directly on it, D if it is in dappled light, N if it is in shade. At the end of the day, count the S marks at each point. That count is the point's direct-sun hours for that date, accurate to about an hour.
A phone helps. Take a photo from the same spot at each check and the photos become the record. If you cannot be home all day, do the morning on one clear day and the afternoon on another clear day within the same week; the sun's path changes little in a week.
Why one day is not enough Go deeper
Physical principle In the northern hemisphere the midday sun is higher in June than in April or September, and it rises and sets further north in summer. A shadow that falls across the bed in March may miss it in June, and a tree that is bare in April is in full leaf by June. Both changes cut the other way too: a neighbor's maple can put a bed in shade for the summer afternoons that mattered most. Convention Oregon State's educator's guide warns that a building may shade the site at critical points in the growing season, which is the same problem.
So a sun map made in early spring is a first measurement, not the answer. Repeat it in midsummer, around late June or July, once trees are in full leaf and the sun is at its highest. If you can only do one, do it as close as possible to the months your fruiting crops will be fruiting. Write both dates on the map.
Phone apps that estimate sun hours from the sun's path can help plan, but they do not see your neighbor's tree or next summer's leaves. Use one to choose the days you check, and the checks themselves for the numbers.
Reading the result
Points on the same bed often differ by two or three hours. That is useful: put tall fruiting crops at the high-count end and greens at the low-count end. Record the lowest point's count as the bed's sun hours in the garden plan's bed sheet, with the date, because the whole bed has to support what you plant there. If no point reaches six hours, look for another spot or for containers you can move before deciding against fruiting crops.
Microclimates: cold air, heat and wind Go deeper
A microclimate is the climate of a small area, such as one corner of a yard, that differs from the climate around it. Two spots in the same yard can differ by several degrees on a still night, and by more on a sunny afternoon. The published frost dates and zones you will look up later describe a weather station, not your bed, so it pays to know which way your spot leans.
Cold air settles low. Physical principle Cold air is denser than warm air, so on a clear, calm night it drains downhill and pools in low ground. Convention Oregon State's guide warns that cool air and frost settle at the base of a slope, especially where a fence holds them in. The USDA's guide to its hardiness map describes the same effect at a smaller scale: a low spot where cold air pools first, and frost pockets, low places among small hills and valleys where frost forms earlier in fall and later in spring than on the ground around them. A garden in a frost pocket may get frost after the published last frost date and before the published first fall frost.
Paving, walls and buildings hold heat. The USDA's map guide names small heat islands caused by blacktop and concrete as a microclimate the map cannot show. A bed against a south- or west-facing wall gets reflected light and stored heat, which helps in spring and can be harsh in July. Convention Oregon State suggests using windows and other reflective surfaces to send light into a shady garden. Raised beds are their own small microclimate: Oregon State notes that in spring they warm and drain faster than the ground.
Wind dries and chills. Oregon State lists windy sites among those to avoid, and suggests building or growing a windbreak where a windy spot is the only choice. Wind matters most for tall crops, young transplants and containers on balconies and roofs.
Write what you notice on the sun map: the low corner where dew lingers, the wall that is warm to the touch at sunset, the gap between houses the wind comes through. These notes explain later surprises, such as frost damage on one bed and not the other.
What a hardiness zone measures, and what it does not
Seed packets, nursery tags and garden websites lean on the hardiness zone. It is useful for some decisions and nearly irrelevant to others, and the line between them is clear once you know what the zone measures.
A hardiness zone is a band on the USDA Plant Hardiness Zone Map defined by the average annual extreme minimum winter temperature: the coldest night of each year, averaged over many years. The USDA's guide describes zones 10 °F wide, split into half zones (such as 7a and 7b) 5 °F wide. The 2023 map, released on November 15, 2023, uses 1991 to 2020 weather data from 13,412 stations. The USDA reports that about half the country shifted to the next warmer half zone compared with the 2012 map, by somewhere between 0 and 5 °F, and says those changes are not necessarily a measure of climate change.
That definition tells you what a zone is for: whether a plant that lives through the winter (a perennial, a shrub, a tree, or garlic overwintering in the ground) can survive the coldest night your area usually gets. Convention University of Illinois Extension makes the point for vegetables directly: most vegetable plants are annuals, so the zones mostly apply to landscape plants. An annual tomato never meets your coldest winter night; it is planted after spring frost and dies at fall frost.
The USDA's own guide lists what the map leaves out: wind, soil type and moisture, humidity, snow cover, winter sunshine, summer heat, and when the cold comes (an early, extreme cold snap can injure plants that would survive the same temperature in January). None of those is in the zone number. Neither is the date of your last spring frost. Two places with the same zone can have very different springs, and two places with the same last frost date can be in different zones.
A zone does not tell anyone when to plant tomatoes. Planting dates for annual vegetables come from frost dates, soil temperature and days to maturity, as Illinois Extension puts it. The zone matters in this course only for crops that overwinter, such as garlic planted in fall (Module 12), and for any perennial you add. If a seed packet's chart is drawn by zone, read the frost dates for your own location instead.
To find your zone, enter your ZIP code on the USDA map site linked below and record it in setup. It takes a minute, and it is worth having for garlic. Then set it aside for the rest of the vegetable season.
Frost dates are probabilities
The dates that run a vegetable garden are the last spring frost and the first fall frost. Most planting guides, including the Colorado State University guide this course uses, count spring planting windows from the "average last spring frost," and fall planting windows backward from the first fall frost. Those dates look precise, printed as a single day. They are not. Each one is a probability read off decades of weather records, and choosing which probability to use is a decision about risk.
The thresholds
NC State Extension's guide to the national freeze tables defines three thresholds used in those tables. Frost in the tables is associated with an air temperature of 36 °F. Physical principle On a clear, still night, leaves can lose heat and frost over while the air at thermometer height is a few degrees warmer. A freeze is air temperature below 32 °F. A hard freeze is below 28 °F. Most published "frost dates" for gardeners, including NOAA's and Purdue's maps, are 32 °F dates. When you look up a date, write down which threshold it uses.
Those thresholds connect to the crops. Colorado State groups vegetables by frost tolerance: hardy crops (pea, lettuce, spinach, cabbage) may survive a frosty nip; semi-hardy crops are less tolerant of a frosty night; tender crops (bush bean, cucumber, summer squash, sweet corn) are intolerant of frost; and very tender crops (tomato, pepper, eggplant, winter squash) are intolerant of frost and of cool spring winds and need daytime temperatures above 60 °F. A 34 °F night with frost on the leaves is a problem for a basil plant and a non-event for spinach.
Reading a probability date
Freeze tables give each threshold at several probabilities. For spring, the date is the last day of spring a temperature at or below the threshold occurs, and the percentage is the chance that it still occurs on or after that date. For fall, the date is the first freeze, and the percentage is the chance it occurs on or before that date. NC State translates a 10 percent chance as about 1 year in 10, and a 90 percent chance as about 9 years in 10.
The 50 percent date is the median: in half of past years the last spring freeze came after it, in half before. Convention When a source says "average last frost date," it almost always means the 50 percent date. NOAA's Climate.gov map of the average last spring freeze defines its dates as the day the chance of a freeze drops below 50 percent, from the 1991 to 2020 Climate Normals, and notes that the Normals also give the dates when the chance drops below 40, 30, 20 and 10 percent. The 10 percent date is later in spring (and earlier in fall): only 1 year in 10 has a freeze beyond it.
| Probability | Spring date means | Fall date means | Risk if you plant frost-tender crops by it |
|---|---|---|---|
| 90 percent | A freeze on or after this date in about 9 years of 10 | A first freeze on or before this date in about 9 years of 10 (a late date) | Aggressive: high risk |
| 50 percent | The median: half of years freeze after it | The median: half of years freeze before it | Moderate |
| 10 percent | A freeze on or after this date in about 1 year of 10 | A first freeze on or before this date in about 1 year of 10 (an early date) | Conservative: low risk |
How far apart are the 50 percent and 10 percent dates? It varies by place. Convention Purdue Extension's Indiana planting calendar maps the 10 percent spring date and notes that about two weeks earlier the chance rises to 50 percent; in fall, it maps the 10 percent date and notes that about two weeks later the chance rises to 50 percent. Two weeks is Indiana's gap, from Indiana's stations. Your table may show a different gap; read it rather than assuming.
The freeze-free period (or frost-free season) is the number of days between the last spring freeze and the first fall freeze. It is the window warm-season crops must fit into, and Module 6 counts back from its end to find the latest planting date for each crop. NC State notes that for this period the risk labels flip: a 90 percent freeze-free period is the conservative figure.
A published frost date is the date by which half of past years had their last 32 °F night. It is not a safe date. The 10 percent date is the lower-risk choice for planting frost-tender crops, and how much risk to accept is yours to choose: a covered seedling you can replace is a different bet from a dozen transplants you started in February. Neither date forecasts this year. NOAA's NCEI notes that each year's actual conditions can vary widely, so the ten-day forecast and a row cover still matter in May. Whatever you choose, write down which probability you used and where it came from.
Where to look up your dates
The national source is NOAA's National Centers for Environmental Information (NCEI), which publishes the 1991 to 2020 U.S. Climate Normals. The Normals include freeze probabilities for individual weather stations at the 32 °F threshold and others. NCEI's page on the last spring freeze links to the Normals and to an interactive map. Station tables are the most detailed source; maps are easier to read but show only the 50 percent date. Note that NCEI's map leaves some places gray because they are too warm, or too cold, to compute a date; see the next section if you are in one.
Your state's extension service is the other source, and often the easier one. Many publish frost-date tables or maps for their own stations: Purdue's Indiana calendar has spring and fall frost maps, and Illinois Extension's planting page gives county dates as "out of 10 years" ranges for 32 °F and 28 °F. Some extension publications still use the older 1981 to 2010 Normals (NC State's freeze guide is one), so note the period. The Extension Foundation's directory finds your state's service; a county extension office or Master Gardener help line can also tell you which station best represents your area.
- Pick a station. Choose the nearest station with a similar elevation and setting. A station on a hilltop airport and a garden in a river valley can have different dates.
- Read four numbers. The 50 percent and 10 percent dates for the last spring 32 °F freeze, and the same two for the first fall freeze.
- Write the source. Station name, threshold, data period (for example 1991 to 2020) and the publisher.
- Choose. Decide which spring date and which fall date you will plan around, and write the reason. Enter those dates in setup; the planting calendar counts from them.
- Go deeper Adjust for your spot. If your sun map notes say frost pocket, lean toward the later spring and earlier fall date.
A short season in numbers. The Lindqvist garden (fictional dates, not a lookup for any real place) has a last spring frost of May 25 and a first fall frost of September 15: 113 days between them. That number alone tells them that every warm-season crop must fit 113 days, before any allowance for harvest or for fall's cooler, slower growth. Module 6 shows that a 100-day winter squash does not fit from direct seed once those allowances are added, and how they settle it. Known: the two dates and the 113 days. Assumed: that the dates they hold are the 50 percent dates (they will confirm which probability with their county extension office, which settles it). Open: the 10 percent dates for the same station, which the station's freeze table settles.
Where frost is rare, heat sets the season
In some places the frost-date method has nothing to count from. NOAA's Climate.gov map of the last spring freeze leaves locations gray where no date could be calculated, and one of the two reasons it gives is warmth: places where the chance of freezing is below 50 percent all spring or all year. NCEI's freeze page puts Puerto Rico in the "too warm to compute" group, a place that very rarely, if ever, sees freezing temperatures.
In those places the calendar turns around. Frost stops being the limit and summer heat becomes it. Cool-season crops that elsewhere go in early spring are grown through the cooler months instead, and the hottest part of summer can be the hardest time to start or grow many crops.
Physical principle Germination has a ceiling as well as a floor. The laboratory data by J. F. Harrington that Oregon State University Extension publishes give maximum soil temperatures for germination: 85 °F for spinach and pea, 95 °F for lettuce and radish (with lettuce's optimum range topping out at 80 °F), but 105 °F for cucumber. In soil hotter than a crop's maximum, its seed does not germinate in that test. Summer soil in a hot climate can pass those figures, which is one reason cool-season crops are sown in fall there. Heat affects more than germination, but this course has no source we trust for each crop's heat limits on flowering and fruit, and we do not guess at them.
So the rarely-frosting gardener needs a different calendar, and the authority for it is local. Ask the authority Use the vegetable planting calendar published by your own state or territory's extension service, which is built for that climate's heat as well as its rare frosts. As an example of what a regional calendar looks like, Virginia Cooperative Extension's planting guide (426-331) gives separate windows for each of Virginia's regions; calendars for warm climates differ much more from the frost-based pattern. The course's calendar respects this: when setup says your area rarely or never frosts, it computes no dates, and shows the crops' frost classes, soil temperatures and days to maturity for use with your local calendar.
Two things still apply in a rarely-frosting climate. The rare frost still happens: a tender crop out on the one cold night of the year is lost, so a forecast and a sheet on hand still matter. And the sun map and microclimate notes matter more, because shade in a hot afternoon can be an asset for lettuce and a west wall can be an oven.
2. See it See it
Both worked examples use the course's fictional cases. Their sun hours, frost dates and climates are invented for the examples; they are not lookups for any real place.
The Okafor sun map and frost dates. The Okafors made a sun map of their two raised beds on a clear day in May, checking five points per bed every hour. Every point first had direct sun at about 10 am and last had it at about 5 pm: 7 hours of direct sun on both beds. They entered 7 hours for each bed in the bed sheet, with the date.
Their notes flagged a maple near the east bed. When they repeated the map in July, the maple in full leaf shaded the east bed from about 4 pm on. That bed now gets less sun in July than in May, and the east bed is where their four tomato plants are. They kept the tomatoes there for this season and wrote the change on the bed sheet and in their notebook, so the harvest log can be read against it.
Their frost dates (fictional) are a 50 percent last spring frost of April 20 and a first fall frost of October 20. Tomato and pepper are very tender, so Colorado State's window starts 2 or more weeks after the average last frost: May 4 at the earliest. Lettuce is hardy, so Colorado State's window is 2 to 4 weeks before: March 23 to April 6.
- Known: 7 hours of direct sun on both beds in May, measured; maple shade on the east bed after about 4 pm in July, measured; a 50 percent last spring frost of April 20 and a first fall frost of October 20.
- Assumed: that the May 4 start for tomatoes, two weeks past the 50 percent date, carries enough margin (Colorado State's convention, accepted by the Okafors); that the station they used represents their yard (their assumption, since the beds sit on flat ground with no frost pocket noted).
- Open: whether the maple shade cuts the tomato harvest, settled by their own harvest log at season end; the 10 percent spring date for the same station, settled by the station's freeze table, which they plan to read before buying transplants.
Marisol's balcony: afternoon sun, rare frost. Marisol Vega's balcony faces west. Her sun map was short: no direct sun until about 1 pm, then sun until about 6 pm, 5 hours. The west wall behind the containers is hot to the touch by late afternoon in summer, and summer afternoons on the balcony are hot. Her area rarely frosts, so the frost-date method gives her nothing to count from and the course's calendar computes no dates for her. She uses her extension service's planting calendar instead.
Five hours meets Oregon State's 4 to 6 for lettuce and radish and falls below its 6 to 8 for fruiting crops. She grows the pepper and the cherry tomato anyway, in the two 10-gallon containers at the brightest end, and treats the result as a test. She moves the lettuce to the shadier end in summer, away from the hot wall, and logs the result.
- Known: 5 hours of direct afternoon sun, measured; a west wall that stores heat; an area that rarely frosts.
- Assumed: that her extension calendar's windows fit her balcony, which is hotter in the afternoon than the open ground the calendar describes (her assumption; she will log sowing dates and results to check it).
- Open: whether 5 hours of afternoon sun is enough for the pepper, settled by her own season's harvest log. The balcony's load limit, settled by the building owner she asked in writing.
Free sources that show the ideas on this page:
- USDA Plant Hardiness Zone Map: your zone by ZIP code, with the definition of what a zone measures.
- When to Expect Your Last Spring Freeze (NOAA NCEI): the national freeze-date map and its links to the 1991 to 2020 Climate Normals.
- Interpreting Freeze/Frost Probabilities (NC State Extension AG-873): how to read 90, 50 and 10 percent dates and choose one by risk.
- An educator's guide to vegetable gardening (Oregon State EM 9032): sun hours by crop type and site pitfalls such as frost in low spots.
3. Practice Practice
Minimum: Make one sun map, look up your frost dates for the nearest station, and enter both in setup; skip the microclimate notes, the zone and the checklist. Time: one minute an hour across one clear day (about 15 minutes in all) for the sun map, 30 minutes for the frost-date lookup, and a repeat sun map in midsummer. Materials: paper and pencil or the bed sheet on this site, a phone with a camera and a timer, a tape measure, three to five markers such as stakes or stones, and an internet connection.
- Sketch the spot. Draw the candidate beds, plot or container spot from above, with the house, fences, trees and anything tall within about 50 feet. Mark north.
- Mark check points. Set three to five markers: the corners and the center of each planned bed, or each container position.
- Check every hour. On a clear day, from early morning to evening, look at each point on the hour. Write S (direct sun), D (dappled) or N (shade). Take a photo from the same place each time.
- Count. Count the S marks per point. Write each point's direct-sun hours and the date on the sketch.
- Go deeper Note the microclimate. Mark low spots, walls and paving that hold heat, and wind gaps. Write one line on each.
- Record the bed sun hours. In the bed sheet, enter each bed's lowest point count as its sun hours, with the date. Set the bed sheet's notes to "recheck in midsummer" if trees or buildings are nearby.
- Go deeper Look up your zone. Enter your ZIP code on the USDA map and write the zone. Enter it in setup.
- Look up your frost dates. Following the steps in the frost section, find the 50 and 10 percent dates for the last spring and first fall 32 °F freeze for a representative station, from NCEI or your extension service. Write the station, threshold, data period and source.
- Choose and enter. Choose your planning dates and write why. Enter them in setup. If your area rarely or never frosts, choose that option in setup, find your extension service's planting calendar, and save its link in your notebook.
- Go deeper Tick the start checklist. Mark "frost dates written down" and "sun map done" in the season start checklist, and put the midsummer recheck in your calendar.
4. Your garden Your garden
5. Check your work Check your work
Criteria
- Is each bed's sun figure a count from hourly checks on a dated day, not an estimate?
- Did you record the lowest point per bed, and plan a midsummer recheck where trees or buildings are near?
- Are fruiting crops assigned to the beds with the most direct sun?
- Do your frost dates name the station or map, the threshold (usually 32 °F), the data period and the probability (50 or 10 percent)?
- Did you choose a probability deliberately and write the reason?
- If your area rarely frosts, do you have your extension service's planting calendar saved instead?
- Is your zone written down, with a note that it is for garlic and perennials, not for planting dates?
Worked example
The Okafor record (fictional) meets most of the criteria. Sun hours are counted from hourly checks on a dated May day, the July recheck is done, and the change on the east bed is written down. Fruiting crops are in beds with 7 measured hours. Their frost dates name the probability (50 percent) and the planting rule they used (Colorado State's very tender window). One criterion is not yet met: they have not read the 10 percent spring date, so their choice of risk is not yet deliberate. They have marked it open, with the station table as what settles it, and the date they will read it.
Common mistakes
- Counting bright shade as sun. If the plants cast no sharp shadow, the hour is not direct sun.
- Mapping once, in April. Leaves and the higher summer sun change the map. A spring map overstates the sun under deciduous trees.
- Treating the average date as safe. The 50 percent date has a later freeze in half of past years.
- Mixing thresholds. A 36 °F frost date and a 32 °F freeze date for the same station differ. Compare like with like.
- Planting by zone. A zone describes the coldest winter night. It does not set planting dates for annual vegetables.
- Forcing a frost calendar onto a frost-free climate. If the map shows your area as too warm to compute, the frost method has nothing to count from; use the local extension calendar.
What is still unresolved
This module gives you the light and the frost window; it does not say what the soil allows (Module 3), which growing method fits the space (Module 4) or when each crop goes in (Module 6, which turns your frost dates into a dated calendar). Whether a spot with fewer hours than the guidance will still produce a given crop is settled only by growing it and logging the harvest. This year's actual last frost is settled by this year's weather; keep an eye on the forecast and keep a row cover on hand for the tender crops.