text stringlengths 0 1.65k |
|---|
Modify the satellite image to reflect 17 years of fluvial geomorphological changes. In the top center, sever the prominent upward-pointing U-shaped bend, routing the main reflective tan channel directly across its neck and transforming the abandoned loop into a darker, disconnected oxbow lake. In the upper-left quadran... |
This is a Landsat 5 satellite image of the Ichilo River in Bolivia (16.80°S, 64.80°W) captured in 1992. The image shows a section of a meandering tropical river flowing through the Bolivian lowlands of the Amazon basin, characterized by active lateral migration and frequent channel avulsion in its alluvial floodplain. ... |
Modify the river's path to depict a meander cutoff by creating a new, thick, light-tan main channel that cuts straight across the narrow neck of land, connecting the river segment in the middle-left directly to the segment in the top-center. Sever the large meandering loop on the right side of the image from the main f... |
This is a Landsat 5 satellite image of the Ichilo River in Bolivia (16.80°S, 64.80°W) captured in 1992. The image shows a section of a meandering tropical river flowing through the Bolivian lowlands of the Amazon basin, characterized by active lateral migration and frequent channel avulsion in its alluvial floodplain. ... |
Alter the river's course to depict meander migration and cutoff. Shift the bend on the left side significantly further to the left, tightening it into a sharp, narrow hairpin loop. Create a new active channel that cuts across the neck of the large rightward meander, connecting the new leftward loop directly to the uppe... |
This is a Landsat 5 satellite image of the Ichilo River in Bolivia (16.80°S, 64.80°W) captured in 1992. The image shows a section of a meandering tropical river flowing through the Bolivian lowlands of the Amazon basin, characterized by active lateral migration and frequent channel avulsion in its alluvial floodplain. ... |
Update the river's course to show a meander cutoff by routing the wide, muddy-brown main channel directly downwards along the left side, cutting through the narrow neck to bypass the large central loop. Convert the bypassed S-shaped loop into an oxbow lake by drastically thinning its channel, disconnecting its ends fro... |
This is a Landsat 5 satellite image of the Ichilo River in Bolivia (16.80°S, 64.80°W) captured in 1992. The image shows a section of a meandering tropical river flowing through the Bolivian lowlands of the Amazon basin, characterized by active lateral migration and frequent channel avulsion in its alluvial floodplain. ... |
Transform the satellite image to reflect 25 years of landscape and infrastructure evolution. Update the main river channel by consolidating the scattered sandbars in the upper right into a large, distinct white sandbar along the eastern bank, making the main water body wider and more uniform in its murky green color. I... |
This is a Landsat 5 satellite image of the Irrawaddy River in Myanmar (21.88°N, 95.98°E) captured in 1996. The image shows a section of Myanmar's largest river, meandering through the central dry zone with active channel migration, point bar accretion, and scroll bar development on its expansive alluvial floodplain. Ba... |
Update the river's course to reflect decades of meander migration and cutoff development. Shift the main channel's bends outward, increasing the overall sinuosity and tightening the necks of the prominent loops in the bottom left and bottom right. In the upper right, create a major cutoff by severing the large westward... |
This is a Landsat 5 satellite image of the Juruá River in Brazil (6.66°S, 69.87°W) captured in 1996. The image shows a section of one of the most sinuous rivers in the Amazon basin, renowned for its exceptionally tight meander bends, frequent neck cutoffs, and abundant oxbow lakes formed throughout its rainforest flood... |
Simulate 28 years of fluvial geomorphological changes by executing a meander cutoff and channel migration. Breach the narrow neck near the center-left, routing the bright tan active river channel directly across it to bypass the large upper-left loop. Convert this bypassed upper-left loop into an oxbow lake by changing... |
This is a Landsat 5 satellite image of the Juruá River in Brazil (6.66°S, 69.87°W) captured in 1996. The image shows a section of one of the most sinuous rivers in the Amazon basin, renowned for its exceptionally tight meander bends, frequent neck cutoffs, and abundant oxbow lakes formed throughout its rainforest flood... |
Update the tight meander bend on the right side of the image to depict a neck cutoff. Create a direct, continuous channel of bright, sandy-colored water straight across the narrow neck, bypassing the loop entirely. Convert the abandoned loop into an oxbow lake by changing its color to a darker, murky green-brown, and v... |
This is a Landsat 5 satellite image of the Juruá River in Brazil (6.66°S, 69.87°W) captured in 1996. The image shows a section of one of the most sinuous rivers in the Amazon basin, renowned for its exceptionally tight meander bends, frequent neck cutoffs, and abundant oxbow lakes formed throughout its rainforest flood... |
Modify the river's path by cutting off the large, tight U-shaped meander loop on the left side. Connect the main channel directly across the narrow neck in the upper left quadrant, creating a continuous, gently curving path that flows downwards to the bottom right. Convert the severed U-shaped loop into an oxbow lake b... |
This is a Landsat 5 satellite image of the Juruá River in Brazil (6.66°S, 69.87°W) captured in 1996. The image shows a section of one of the most sinuous rivers in the Amazon basin, renowned for its exceptionally tight meander bends, frequent neck cutoffs, and abundant oxbow lakes formed throughout its rainforest flood... |
Modify the bright river channel to depict meander migration and bend tightening. Extend the prominent leftmost U-shaped bend further to the left. Shift the upper arm of this bend downward and the lower arm upward, significantly narrowing the neck of land between them. Move the river segment exiting toward the right edg... |
This is a Landsat 5 satellite image of the Juruá River in Brazil (6.66°S, 69.87°W) captured in 1996. The image shows a section of one of the most sinuous rivers in the Amazon basin, renowned for its exceptionally tight meander bends, frequent neck cutoffs, and abundant oxbow lakes formed throughout its rainforest flood... |
Modify the river's course to show significant meander cutoffs and channel straightening in the central section. Starting from the upper-middle, bypass the sharp left-pointing meander loop, routing the main sandy-brown channel straight south and converting the abandoned loop into a dark green oxbow lake. Immediately bel... |
This is a Landsat 5 satellite image of the Mamoré River in Bolivia (14.88°S, 65.04°W) captured in 1988. The image shows a section of a major tributary of the Madeira River, meandering through the Bolivian Beni floodplain with pronounced meander loops, active bank erosion, and a dynamic network of abandoned channel trac... |
Modify the river channel to depict meander cutoff and lateral migration. In the left-center, locate the tight meander loop with a narrow neck. Cut a new, thick, bright sandy channel straight across this neck, connecting the upstream and downstream reaches directly to bypass the large southern loop. Sever this bypassed ... |
This is a Landsat 7 satellite image of the Purus River in Brazil (7.30°S, 64.85°W) captured in 2001. The image shows a section of a highly sinuous Amazon tributary known for its exceptionally tight meander bends, rapid lateral migration rates, and prolific oxbow lake formation across its broad tropical floodplain. Base... |
Update the satellite image to reflect 23 years of fluvial geomorphology changes. Change the color of the main river from a uniform orange-tan to a murky greenish-brown, and add bright white and pale yellow sandy point bars along the inner edges of the curves. Breach the narrow neck of land on the right side of the larg... |
This is a Landsat 7 satellite image of the Purus River in Brazil (7.30°S, 64.85°W) captured in 2001. The image shows a section of a highly sinuous Amazon tributary known for its exceptionally tight meander bends, rapid lateral migration rates, and prolific oxbow lake formation across its broad tropical floodplain. Base... |
Replace the side-by-side layout with a single full-frame satellite image of the estuary. Change the water throughout the bay and river channels to a murky, opaque green. Significantly expand the reddish-brown sandbars and sediment flats within the river delta, narrowing the green water channels that weave between them.... |
The two side-by-side Landsat-5 satellite images show the Betsiboka River Estuary near Mahajanga, Madagascar (15.88°S, 46.30°E). The left image is from 1990 and the right image from 1995. The Betsiboka is one of the world's most sediment-laden rivers due to extreme upstream deforestation and laterite soil erosion, and t... |
Convert the side-by-side image into a single aerial view of the river delta, matching the spatial extent of one panel. Remove the milky turbidity plumes from the ocean, rendering the water a clear, uniform deep blue. At the river mouth, build a massive, solid light-gray sediment delta that extends prominently into the ... |
The two side-by-side high-resolution NAIP aerial images show the Elwha River Delta at the Strait of Juan de Fuca, Washington, USA (48.15°N, 123.57°W). The left image is from 2011 (just before the start of dam removal) and the right image from 2013 (during active dam removal). The Elwha and Glines Canyon dams trapped ov... |
Transform the side-by-side satellite image into a single, unified frame showing the same delta region. Update the land cover to reflect widespread deforestation by replacing extensive areas of dark green mangroves with lighter brown and tan agricultural plots across the central peninsula and the flanking landmasses to ... |
The two side-by-side Landsat-5 satellite images show a portion of the Irrawaddy (Ayeyarwady) River Delta in Myanmar (15.87°N, 95.20°E). The left image is from 1990 and the right image from 1995. The Irrawaddy Delta is one of Asia's largest tide-influenced deltas, and these two snapshots capture the initial changes in l... |
Replace the side-by-side layout with a single, unified satellite image of the delta showing advanced progradation. Remove the vertical split and present a continuous view where the central river entering from the top branches into a massive, radial network of sharply defined, thick brown distributary channels extending... |
The two side-by-side satellite images show the Wax Lake Delta in Atchafalaya Bay, Louisiana, USA (29.55°N, 91.45°W). The left image was captured by Landsat-5 in 1990 and the right image by Landsat-5 in 1995. Together they reveal the early growth trajectory of this river-dominated delta following the 1973 opening of the... |
Replace the side-by-side split image with a single, full-width satellite view of the river delta. Update the landscape so the main river channel flows eastward and then curves sharply to the northeast, extending a massive, newly formed brown sediment lobe far into the ocean. Fill the water surrounding this new northeas... |
The two side-by-side Sentinel-2 satellite images show the Yellow River (Huanghe) Delta near Dongying, Shandong, China (37.76°N, 119.15°E). The left image is from 2016 and the right image from 2018. The Yellow River delta is one of the most dynamic deltas on Earth, and these two snapshots capture the ongoing channel mig... |
Replace the large, jagged black void in the upper-middle section of the river estuary with flowing, murky green water and reddish-brown sediment streaks that seamlessly connect the surrounding channels. Fill in the scattered black artifact spots across the landmasses on the left and bottom with the surrounding brown, r... |
This Landsat satellite image shows the Betsiboka River Estuary near Mahajanga, Madagascar (15.88°S, 46.30°E) captured in 1990. The Betsiboka is one of the most sediment-laden rivers in the world due to extreme deforestation and laterite soil erosion in its upstream catchment. The estuary shows a complex braided channel... |
Modify the satellite image to show significant sediment accumulation at the river delta. At the river mouth where it meets the ocean, expand the landmass outward into the dark blue water by adding a large, light-grey deposit of sand and gravel. This new sediment should form a prominent, rounded bulge that extends the c... |
This high-resolution satellite image shows the Elwha River Delta at the Strait of Juan de Fuca, Washington, USA (48.15°N, 123.57°W) captured in 2011, at the start of the largest dam removal project in U.S. history. The Elwha and Glines Canyon dams had trapped over 30 million cubic yards of sediment for nearly a century... |
Replace large areas of the dark green mangrove forests across all the main landmasses with patchy, lighter brown and tan agricultural fields, leaving fragmented dark green vegetation primarily along the edges of the intricate tidal creeks. Enlarge the small island in the lower-left quadrant, making it slightly broader.... |
This Landsat satellite image shows a portion of the Irrawaddy (Ayeyarwady) River Delta in Myanmar (15.87°N, 95.20°E) captured in 1990. The Irrawaddy Delta is one of Asia's largest tide-influenced deltas, featuring a complex network of distributary channels, tidal creeks, and extensive mangrove forests. In 1990, the del... |
Expand the river delta significantly further into the bay by extending the brown, branching distributary channels outward and downward. Fill the spaces between the upper and middle sections of these lengthened channels with solid, dark green landmass to represent established vegetation. Increase the overall contrast, s... |
This satellite image shows the Wax Lake Delta in Atchafalaya Bay, Louisiana, USA (29.55°N, 91.45°W) as captured by Landsat in 1990. At this time the delta is in an early growth stage following the 1973 opening of the Wax Lake Outlet — a relatively small, nascent fan of sediment is being deposited into the shallow bay. ... |
Transform the main river channel flowing from the left into a thick, opaque, light brown band to represent a heavy sediment load. Significantly expand the active northeastern delta lobe by extending a massive, fan-shaped deposit of light brown sediment far into the sea, replacing the previously defined channels with a ... |
This Sentinel-2 satellite image shows the Yellow River (Huanghe) Delta near Dongying, Shandong, China (37.76°N, 119.15°E) captured in 2016. The Yellow River delta is one of the most dynamic deltas on Earth, characterized by extremely high sediment loads and repeated engineered channel avulsions. In 2016, the active dis... |
Shift all sand dunes, ridges, and surface ripples slightly toward the bottom-left to simulate wind-driven migration. |
This is a satellite image of Bodélé Depression Site A (16.85°N, 17.35°E) captured in January 2023. The scene shows the northern Bodélé Depression, one of the world's most prolific dust source regions. This area sits in a topographic gap between the Tibesti and Ennedi mountain ranges, where wind-tunnel acceleration (the... |
Based on the known aeolian geomorphic processes at this site — including wind-tunnel acceleration through the topographic gap, yardang erosion, deflation of diatomite substrates, and barchan dune migration driven by the northeasterly Harmattan winds — generate the expected appearance of this scene 3 years later in Janu... |
Shift all crescent-shaped barchan dunes and the textured sand formations in the upper left diagonally toward the bottom-left corner to simulate wind-driven migration. Preserve the exact size, shape, orientation, and shading of every dune, ensuring their edges blend seamlessly into the smooth, flat, light-orange sandy b... |
This is a satellite image of Bodélé Depression Site B (16.70°N, 17.10°E) captured in January 2023. The scene shows the central Bodélé Depression, where intense deflation of Mega-Chad paleo-lake diatomite deposits creates a dense yardang field. The wind-funnel effect between the Tibesti Mountains to the north and the En... |
Based on the known aeolian geomorphic processes at this site — including wind-tunnel acceleration through the topographic gap, yardang erosion, deflation of diatomite substrates, and barchan dune migration driven by the northeasterly Harmattan winds — generate the expected appearance of this scene 3 years later in Janu... |
Shift all the prominent, lighter-orange sand dunes slightly towards the bottom-left corner of the image to simulate their migration over a three-year period driven by northeasterly winds. Maintain the shapes and relative sizes of the dunes, and ensure the underlying lighter-colored, textured terrain and the overall sat... |
This is a satellite image of Bodélé Depression Site D (16.60°N, 16.90°E) captured in January 2023. The scene shows the southwestern Bodélé Depression, near the narrowest section of the topographic gap where wind-tunnel acceleration is most intense. This site shows prominent linear yardang ridges carved from friable dia... |
Based on the known aeolian geomorphic processes at this site — including wind-tunnel acceleration through the topographic gap, yardang erosion, deflation of diatomite substrates, and barchan dune migration driven by the northeasterly Harmattan winds — generate the expected appearance of this scene 3 years later in Janu... |
Shift the entire field of barchan dunes in the lower two-thirds of the image slightly toward the bottom-left (southwest) to simulate three years of wind-driven migration. Maintain the exact shapes, shading, and relative positions of the dunes as they move. Keep the lighter, textured terrain and white patches in the upp... |
This is a satellite image of Bodélé Depression Site F (16.75°N, 17.50°E) captured in January 2023. The scene shows the eastern margin of the Bodélé Depression, where the wind corridor begins to widen as it exits the Tibesti–Ennedi gap. This transitional zone shows a mix of degraded yardang remnants and downwind barchan... |
Based on the known aeolian geomorphic processes at this site — including wind-tunnel acceleration through the topographic gap, yardang erosion, deflation of diatomite substrates, and barchan dune migration driven by the northeasterly Harmattan winds — generate the expected appearance of this scene 3 years later in Janu... |
Shift all the distinct, golden-tan, crescent-shaped sand dunes diagonally towards the bottom-left corner by approximately one-tenth of the image width. This translation applies to the prominent central dune, the cluster in the upper right, the dune on the left edge, and all smaller dunes scattered across the scene. Kee... |
This is a satellite image of Bodélé Depression Site G (16.55°N, 16.80°E) captured in January 2023. The scene shows the southwesternmost sector of the Bodélé Depression, positioned at the throat of the wind funnel. This site experiences the highest sustained wind speeds in the region and displays closely spaced, deeply ... |
Based on the known aeolian geomorphic processes at this site — including wind-tunnel acceleration through the topographic gap, yardang erosion, deflation of diatomite substrates, and barchan dune migration driven by the northeasterly Harmattan winds — generate the expected appearance of this scene 3 years later in Janu... |
Shift all the crescent-shaped sand dunes and their associated highlights and shadows slightly toward the bottom-left corner of the image to simulate downwind migration. Keep the underlying terrain, overall spatial extent, and the orange-tan color palette and texture of the satellite imagery unchanged. |
This is a satellite image of Chad Barchan Dune Field (16.50°N, 18.50°E) captured in January 2023. The scene shows a barchan dune field east of the Bodélé Depression in central Chad. These crescent-shaped dunes migrate downwind under the influence of the persistent northeasterly Harmattan winds. The dune spacing, horn o... |
Based on the known aeolian geomorphic processes at this site — including wind-tunnel acceleration through the topographic gap, yardang erosion, deflation of diatomite substrates, and barchan dune migration driven by the northeasterly Harmattan winds — generate the expected appearance of this scene 3 years later in Janu... |
Draw a large, curved directional arrow to illustrate the net sand transport direction. The arrow's path must originate on the upper right, sunlit slope of the dune, arch smoothly over the sharp crest line, and terminate with its arrowhead pointing downwards and to the left into the dark, shadowed slip face. Style the a... |
This photograph shows a dune with a clearly expressed crest and lee slip face. Based on the dune form, infer the net sand transport direction and draw one directional arrow showing the movement path of the migrating sand. Keep the dune, sky, ground, moon, and vegetation unchanged. |
Draw a large, curved arrow indicating the direction of sand movement across the dune. The arrow should start on the right, sunlit, gently sloping side of the dune, arc upwards and to the left over the crest, and point downwards onto the darker, shadowed, steep slip face on the left. The arrow must have a thick bright c... |
This photograph shows a dune with an asymmetric profile, a crest, and a steep slip face. Based on the dune form, infer the net sand transport direction and draw one directional arrow showing the movement path of the migrating sand. Keep the dune, sky, ground, and vegetation unchanged. |
Add two large, curved directional arrows pointing downwards and to the right to indicate the sand transport direction. Place the top arrow over the upper-left slope of the highest background dune, curving down to point at its steep, shadowed slip face. Place the second arrow below it, over the left slope of the promine... |
This photograph shows a dune field with multiple superposed dune forms and visible lee slopes. Based on the dune forms, infer the net sand transport direction and draw directional arrows for the two main migrating dune bodies. Keep the dunes, sky, and ground unchanged. |
Transform the satellite image into a digital elevation model (DEM) visualization using a continuous false-color gradient where deep blue represents the lowest elevations and red represents the highest. Color the broad river valley that spans from the middle-left down to the bottom-right in shades of deep blue and cyan.... |
This satellite image shows the Río Negro, Patagonia area (39.3°S, 65.9°W), Argentina. The river flows generally east through this region. |
Generate the corresponding elevation profile (DEM visualization) for this area. The elevation map should use a blue-to-red color palette (blue = low elevation, red = high elevation) and reveal the cross-river topographic asymmetry at this location. |
This site is in the Southern Hemisphere at 39.3° latitude. |
Maintain the same geographic extent and spatial alignment as the input satellite image. |
Transform the satellite image into a digital elevation model (DEM) visualization using a continuous color gradient. Apply a color palette where low elevations are represented by solid cyan and light blue, transitioning through green and yellow for intermediate elevations, to orange and bright red for the highest terrai... |
This satellite image shows the Paraguay River at Asunción area (25.3°S, 57.6°W), Paraguay. The river flows generally south through this region. |
Generate the corresponding elevation profile (DEM visualization) for this area. The elevation map should use a blue-to-red color palette (blue = low elevation, red = high elevation) and reveal the cross-river topographic asymmetry at this location. |
This site is in the Southern Hemisphere at 25.3° latitude. |
Maintain the same geographic extent and spatial alignment as the input satellite image. |
Transform the satellite image into a digital elevation model (DEM) visualization using a false-color heatmap. Replace the natural terrain colors such that the lowest elevations, specifically the river channels and water bodies, are rendered in dark blue. Color the flat floodplains occupying the left and upper portions ... |
This satellite image shows the Amur River at Khabarovsk area (48.5°N, 134.9°E), Russia. The river flows generally northeast through this region. |
Generate the corresponding elevation profile (DEM visualization) for this area. The elevation map should use a blue-to-red color palette (blue = low elevation, red = high elevation) and reveal the cross-river topographic asymmetry at this location. |
This site is in the Northern Hemisphere at 48.5° latitude. |
Maintain the same geographic extent and spatial alignment as the input satellite image. |
End of preview. Expand in Data Studio
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
GeoR-Bench
GeoR-Bench is an anonymous geoscience image-text benchmark for evaluating multimodal Intelligence on Earth science tasks. It contains 440 instance, where each instance contributes one input+output+prompt. Each instance includes paired input and target images, prompt and judge rubrics spanning satellite imagery, maps, schematics, and other geoscience visual representations.",
Dataset structure
Each benchmark example is stored as a directory containing five canonical files:
input.pngoutput.pnginstruction_spcfic_prompt.txtreason_prompt.txtjudge.md
A global machine-readable manifest is provided at manifest.jsonl, with one row per dataset record.
Metadata files
manifest.jsonl: global example manifestcroissant.json: manual Croissant 1.1 JSON-LD descriptor for submission and external toolingLICENSE: dataset license
Attribution
This release keeps creator and citation metadata anonymous for submission purposes.
License
This dataset is released under CC BY 4.0: https://creativecommons.org/licenses/by/4.0/
- Downloads last month
- 893