Despite cold weather in February, cherry blossoms in Washington D.C. bloomed earlier than expected. The National Park Service declared peak bloom at the tidal basin March 26. My early prediction (ten days ago) was off by four days.
Much like a dashboard GPS device might say … “Recalculating …”
March 26 : Peak bloom at Washington D.C. March 31 : Kwanzan (late-season cherry) begin bloom at Washington D.C. April 7 : Peak bloom at Philadelphia April 15 : Kwanzan (late-season cherry) begin bloom at Philadelphia April 19 : Peak bloom at Boston April 27 : Kwanzan (late-season cherry) begin bloom at Boston
March 1st, I wrote this to a friend: “Temperatures in January & February 2026 were colder than average, which may suggest peak bloom first week of April. However, the National Weather Service is predicting warmer than average temperatures in March, which would accelerate it to end of March.”
March 12 in Washington, the range of predictions regarding peak bloom at the tidal basin suggest March 28 at the earliest and April 7 at the latest. Currently in Washington D.C., according to the National Park Service, the early-blooming Okame Cherry trees were in bloom while the Yoshino Cherry trees at the Tidal Basin had reached the florets visible stage.” While Okame flowers may be affected by frost this week, blossoms that have not yet appeared and will not be affected.
On March 17, both D.C. and Boston will experience a hard freeze as temperature drops below 30º F (-1º C). Furthermore, most of this week, the daily high temperatures in Washington D.C. are very similar to temperatures in Boston Massachusetts. That’s unusual.
Unless the weather in D.C. does something very peculiar in the next 10 days, Yoshino peak bloom should hit between March 30 to April 2.
Boston
In light of how the season is developing in Washington D.C., what does that imply for cherry bloom around Boston? It’s really too early to predict, but … maybe the Okame Cherry begin blooming by mid April and the Yoshino peak bloom around last week of April.
Though it is true that spring weather advances from south to north, the progression is variable from year to year. Some years, spring progress can stall around New York City such that all points north remain on the cold side and flowering trees bloom late. Unfortunately, a prediction of winter’s end from Punxsutawney Phill is useless.
Three recent examples to share, each created using AI tools from Topaz Labs. Though I have long relied upon photo tools from Topaz, the AI features are quite new, added within the past several months.
The first image here shows restore and recolor an old black & white photo. This was done using a Topaz online tool called Mosaic. Each use of the online tools requires some AI credits. Use these enough and your credits will be exhausted, requiring purchasing more credits.
In general, the result is quite impressive but there is a flaw. Look at the middle finger of the man’s right hand – the color is the same as the chair.
The second and third images here show two separate attempts to fix a photo of a cardinal in a cherry tree. I shot this in New York’s Central Park last spring but my camera wasn’t the best and I was too far away.
The first attempt employed Topaz Photo AI (desktop app) and a new AI model for wildlife, which is still in beta test.
The second attempt uses Topaz Gigapixel AI (desktop app) and a new AI model called Redefine Realistic, which is still in Beta test. This model can optionally use a prompt and you can see the prompt I used (“improve details in the red bird”). Not only does the result show detail in the feathers, but now the bird has an eye. However, the bird’s beak isn’t quite right and now has a nostril.
The changes to the bird are a bit heavy-handed and Gigapixel AI did not limit its changes to only the bird; some undesirable artifacts were introduced to the flowers and in the soft out-of-focus background. This implies a bit more work in Photoshop to blend the revised bird image onto the unmodified background image. I might choose to blend some of the different results from Photo AI and Gigapixel AI.
Though Gigapixel is commonly used to increase image resolution, you can see here that I did not upscale the image (Upscale = 1x).
The desktop apps (Photo Studio and Gigapixel AI) allow choosing either cloud computing or local computer resources. I chose to use my local computer, which does have a GPU. The sound of my computer suggested the GPU fans were spinning fast and Windows Task Manager told me that the GPU was running at full capacity. Using cloud computing will consume AI credits but using the local computer does not require any AI credits.
