Press release · National Infrastructure Authority (ANIN) · Communications Unit · 19 September 2025
A large-scale topographic survey will cover approximately 397,155 hectares and 480 kilometers, providing the data needed to design protection works and integrated solutions against flooding and erosion.
The National Infrastructure Authority (ANIN) has begun LiDAR flights over the Piura River basin as part of profile-level pre-investment studies for the project “Improvement and Expansion of Riverbank Protection Services in Areas Vulnerable to Fluvial Flooding and Erosion in the Piura River Basin.”
The campaign, scheduled to continue through October, operates from the airstrip at the populated center of Yapatera. It will cover the upper, middle, and lower basin of the Piura River — an area of approximately 397,155.28 hectares along 480 kilometers, from the headwaters to the river’s outlet to the sea at Laguna La Niña.
Airborne LiDAR is a topographic survey method that uses sensors mounted on aircraft — in this case a specially equipped helicopter — that emit laser pulses toward the Earth’s surface. Those pulses capture millions of georeferenced points at millimeter-level precision and produce detailed digital models of the terrain and its cover.
The survey is expected to deliver:
- Digital elevation models (DEM) and digital terrain models (DTM) that accurately represent the topography of the basin and its tributaries.
- Up-to-date technical information for the pre-investment profile, supporting the design of detention and regulation works, riverbank protection, erosion control, and channelization as part of an integrated package of solutions for Piura.
ANIN’s technical team noted that this is one of the largest geodetic surveys ever carried out in Peru and will provide high-quality inputs for project formulation.
Before operations began, the contractor’s equipment was inspected to confirm that the required technical conditions were in place. The first LiDAR flight was completed successfully.
With this work, ANIN reaffirms its commitment to the Comprehensive Flood Control Project for the Piura River and to the technical, responsible design and delivery of infrastructure that strengthens the safety of Piuran families against climate-related hazards.
Technical note: pulse density, DEM and DTM in hydraulic design
By GeoKapæc
The official release does not publish pulse density or vertical accuracy for this contract. What follows is the usual technical frame for reading — and specifying — LiDAR products for flood modeling. It is not this flight’s specification.
Vertical accuracy (Z)
What sets LiDAR apart from conventional photogrammetry is the quality of the Z coordinate. The sensor times the pulse and assigns a height to each return; with GNSS/IMU, calibration and geodetic control, that Z is expressed in a common height system. The release’s “millimeter-level precision” describes the potential of the laser rangefinder and the control network, not the RMSEz of the finished DTM. For basin and floodplain products, usable vertical accuracy is typically centimetre-level on bare ground (often near 10 cm RMSEz under QL2-class specifications) and degrades under dense vegetation, slopes and water. For detention works and defenses, a 10–20 cm bias on a levee crest or floodplain changes the inundated area. The figure to require in the pre-investment profile is vertical accuracy checked against benchmarks — bare and vegetated ground separately — not the size of the pulse.
The coordinate that justifies a LiDAR flight for hydraulics is Z, not XY. Three different figures get conflated: rangefinder resolution (millimetres), GNSS/IMU trajectory accuracy (usually centimetres), and RMSEz of the delivered DTM, which also includes classification, interpolation and the height model. ASPRS / USGS-style specs split non-vegetated vertical accuracy (NVA) from vegetated vertical accuracy (VVA). Without those two numbers, and without stating whether heights are ellipsoidal or orthometric, a 2D model cannot be audited.
Pulse density is not point density
Pulse density (pulses/m²): shots the sensor emits per square metre. Return / point density (pts/m²): recorded echoes. One pulse can yield several returns (ground, understory, canopy, roof).
For hydraulics the decisive number is how many of those echoes classify as ground. Under a dense canopy or riparian cane, ground density can be a fraction of pulse density.
| Typical use | Order of magnitude |
|---|---|
| Regional / basin relief | ~1–4 pulses/m² |
| Floodplain (QL2-class) | ≥2–8 pts/m² (first return) |
| Banks, levees, works, urban channels | 8–16+ pts/m²; often helicopter at lower AGL |
More pulses do not “fix” a DTM if ground classification is weak, breaklines are missing, or near-infrared energy is lost on water.
DEM, DSM and DTM are not interchangeable
- DSM (digital surface model): first returns. Includes vegetation, roofs and bridges.
- DTM / bare-earth DEM: classified ground only. This is the terrain the hydraulic model uses: water runs on the ground, not on canopies or rooftops.
- Hydro-enforced/flattened DEM in official Peruvian texts is often used loosely. Ask whether the raster is hydro-enforced and whether water is hydro-flattened.
What changes in hydraulic design
- Detention and regulation. Storage volume is computed on the DTM, not the DSM. An 8–12 m canopy treated as “ground” shrinks storage and shifts the hydrograph.
- Channelization and bank protection. Bank crest, toe and berms need high density plus breaklines. A 1–2 m raster with no breaklines rounds the edge and underestimates velocities.
- Erosion control. Roughness (Manning’s n or 2D equivalents) leans on DSM−DTM (vegetation height) and bank shape. Without both, the hydrodynamic model is guessing.
- Water and bathymetry. Near-infrared LiDAR barely penetrates turbid water. The wetted bed of the Piura will need bathymetry (sonar, green LiDAR or cross-sections) if a continuous in-channel / floodplain 1D/2D model is required.
- Bridges and culverts. The DSM “fills” the opening. Without hydro-enforcement (cut the deck, open the barrel) the model treats the crossing as a dam.
How to read the release.
“Millimeter-level precision” describes the potential of the sensor and the geodetic control network, not automatically the accuracy grade of the finished DTM over vegetation or slopes.
A product that is actually usable for the pre-investment profile should declare, at a minimum pulse and point density, accuracy level on bare and vegetated ground, DEM cell size, reference systems, and whether hydro-enforcement and channel bathymetry are included.
The topography stage of the same project was announced separately on 7 September 2025: 10 geodetic reference points and 38 benchmarks (BMS), a budget of S/ 8.9 million within S/ 24 million for project formulation, and a comprehensive project estimated at more than S/ 4.9 billion to protect about 1.5 million people in the provinces of Piura, Morropón, Ayabaca, Huancabamba and Sechura.
Source: Plataforma del Estado Peruano · National Infrastructure Authority (ANIN), 19 September 2025. Original: gob.pe/institucion/anin/noticias/1249148. Formulation context (7 Sep 2025): topographic study for the Flood Control Project. The official lead cites “393 thousand hectares”; the body specifies 397,155.28 ha. The body figure is used here. The technical note is editorial commentary and is not part of the ANIN release.