---
title: "Investigating the Value of Adding Diverse ML Models to the IMPROVER Blend"
ocid: "ocds-h6vhtk-078a0f"
canonical_url: "https://d3tenders.com/contract/?ocid=ocds-h6vhtk-078a0f"
markdown_url: "https://d3tenders.com/contract/ocds-h6vhtk-078a0f.md"
json_url: "https://d3tenders.com/contract/ocds-h6vhtk-078a0f.json"
source: "Find A Tender Service"
current_stage: "Award"
buyer: "Met Office"
published: "2026-10-20"
---

# Investigating the Value of Adding Diverse ML Models to the IMPROVER Blend

Buyer: Met Office  
Current stage: Award  
OCID: ocds-h6vhtk-078a0f

[View canonical contract page](https://d3tenders.com/contract/?ocid=ocds-h6vhtk-078a0f)  
[Download OCDS JSON](https://d3tenders.com/contract/ocds-h6vhtk-078a0f.json)

## Summary

The Met Office commissioned research into whether adding a wider range of machine-learning weather prediction models to its IMPROVER forecast blend improves forecasts enough to justify the added operational complexity. The work examines the value of model diversity, how AI models complement existing numerical weather prediction systems, the marginal benefit of additional models, and which combinations perform best. It is intended to inform operational decisions and future Met Office strategy, not just theoretical research. All meteorological variables are in scope, with particular interest in wind, temperature and precipitation, everyday performance and extreme events, and forecast lead times beyond roughly three to five days. Delivery covers the UK. The service is weather-forecasting research (CPV 73110000).

The procurement is complete and the contract has been awarded to the University of Exeter. The Met Office used a competitive award under a framework, with one lot. The award was made on 7 September 2026; the contract was signed on 24 September 2026 and runs from 16 October 2026 to 31 March 2028. Its value is £297,014.28 excluding VAT (£356,417.13 including VAT). The requirement was for research services, evaluated through the framework route. The contract is active. The award and contract cover the single research requirement described above.

This award shows demand for applied research that tests whether machine-learning weather forecasts improve an operational blend, especially when combined with established numerical weather prediction. A credible competitor when this requirement next comes to market would need expertise in machine-learning weather prediction, forecast blending and quantitative evaluation of model skill and complementarity. The work calls for comparing model combinations, assessing the incremental value of additional models and identifying where gains occur across forecast lead times and meteorological variables. Suppliers should be able to connect measured forecast improvements to operational usefulness, including performance in everyday conditions and extremes over the UK. Experience translating technical findings into evidence that can inform a national weather service’s operational choices and strategy would be directly relevant.

## Notice

The central research question is: Does incorporating a broader and more diverse range of machine learning weather forecasting models into the IMPROVER blend deliver sufficient forecast improvement and operational value to justify the additional complexity of implementation and operation? The Met Office is particularly interested in understanding: * The value of model diversity within forecast blends. * Whether AI models provide additional skill when blended with existing NWP systems. * The marginal benefit of adding successive AI models. Whether a point of diminishing returns exists beyond which additional models provide little operational value. * Which combinations of models deliver the greatest forecast benefit. * Where benefits are largest? For example, at what lead times and for what parameters? * Whether complementary behaviour between models is more important than individual model skill alone. * Whether increasing the number of MLWP members in the blend could compensate for fewer NWP ensemble members in the blend The work should focus on operationally relevant questions rather than purely theoretical investigations. While forecast skill improvements are important, the ultimate goal is to determine whether these improvements are sufficient to influence future operational decisions. The outputs should therefore provide evidence that can support future Met Office strategy. Particular interest exists in forecasting lead times beyond approximately three to five days, where traditional NWP systems begin to lose skill and where MLWP models may provide the greatest opportunity for improvement. However, shorter lead times should not be excluded if evidence demonstrates that complementary model behaviour leads to measurable forecast improvements. All meteorological parameters are of interest and IMPROVER offers the opportunity to blend a different number of models for different parameters. The core variables of wind, temperature and precipitation are of most interest. We are interested in general 'every day' performance as well as extremes, particularly over the UK.

## Key Details

| Field | Value |
| --- | --- |
| Publication source | Find A Tender Service |
| Latest notice | https://www.find-tender.service.gov.uk/Notice/098875-2026 |
| Notice type | UK7 - Contract Details Notice |
| Procurement type | Standard |
| Procurement category | Services |
| Procurement method | Selective |
| Procurement method details | Award under framework |
| Tender suitability | Not specified |
| Awardee scale | Large |
| All stages | Award |

## Dates

| Field | Value |
| --- | --- |
| Publication date | 20 Oct 2026 |
| Submission deadline | Not specified |
| Future notice date | Not specified |
| Award date | 7 Sep 2026 |
| Contract period | 16 Oct 2026 - 31 Mar 2028 |
| Recurrence | Not specified |

## Values

| Field | Value |
| --- | --- |
| Tender value | Not specified |
| Lots value | Not specified |
| Awards value | £297,014 |
| Contracts value | £297,014 |

## Status

| Field | Value |
| --- | --- |
| Tender status | Complete |
| Lots status | Complete |
| Awards status | Active |
| Contracts status | Active |

## Buyer

| Field | Value |
| --- | --- |
| Main buyer | Met Office |
| Locality | EXETER |
| Postcode area | Exeter |
| Postcode | EX1 3PB |
| Country | England |
| ITL 1 | TLK South West (England) |
| ITL 2 | TLK4 Devon |
| ITL 3 | TLK43 Devon CC |
| Local authority | Exeter |
| Electoral ward | Pinhoe |
| Westminster constituency | Exmouth and Exeter East |
| Delivery location | Not specified |

## Supplier

| Field | Value |
| --- | --- |
| Number of suppliers | 1 |
| Supplier names | University of Exeter |

## CPV Codes

### Divisions

- 73 - Research and development services and related consultancy services

### Codes

- 73110000 - Research services

## Release History

- 20 Oct 2026 at 11:13 - Award - UK7 - Contract Details Notice - https://www.find-tender.service.gov.uk/Notice/098875-2026
- 16 Oct 2026 at 16:26 - Award - UK6 - Contract Award Notice - https://www.find-tender.service.gov.uk/Notice/098166-2026

## Documents

- https://www.find-tender.service.gov.uk/Notice/098875-2026
  20th October 2026 - Contract details notice on Find a Tender
- https://www.find-tender.service.gov.uk/Notice/098166-2026
  16th October 2026 - Contract award notice on Find a Tender
- https://www.find-tender.service.gov.uk/Notice/Attachment/A-24380
  Redacted contract

## Notice URLs

- http://www.exeter.ac.uk
- https://www.find-tender.service.gov.uk/api/1.0/ocdsReleasePackages/ocds-h6vhtk-050b9a
- https://www.legislation.gov.uk/ukpga/2023/54/contents
- https://www.metoffice.gov.uk/

## Provenance

This Markdown file is an alternate public rendering of the D3 Tenders contract record. The canonical page is https://d3tenders.com/contract/?ocid=ocds-h6vhtk-078a0f. The underlying structured data is available as OCDS JSON at https://d3tenders.com/contract/ocds-h6vhtk-078a0f.json.