Mujāhid and Ṭāwūs, from Ibn ʿAbbās.
I remark: This chain of transmission is weak, usable only for corroboration. Al-Mundhirī said in al-Targhīb (1/271):
“Its isnād is close to ḥasan and it has supporting narrations.”
I further say: It seems to me that the report ultimately goes back to Buraydah, for I located another partial route for it. Al-Ṭabarānī related it in al-Awsat (1/85/1, in the combined edition with al-Ṣaghīr) and Tammām in al-Fawāʾid (folio 148-149) through Marwān ibn Muḥammad al-Ṭaṭarī, who said: “Sulaymān ibn Mūsā Abū Dāwūd al-Kūfī narrated to us, from Faḍl ibn Marzūq (while al-Fawāʾid reads: Faḍl ibn Ghazwān), from ʿAbd Allāh ibn Buraydah, from his father, in marfūʿ form:
‘No people ever withhold the zakāh without Allah testing them with years of drought.’ ”
Al-Ṭabarānī then commented: “No one relates it except Sulaymān, and Marwān alone transmits it from him.”
I say: Marwān is reliable. Sulaymān ibn Mūsā Abū Dāwūd al-Kūfī is “decent,” as al-Dhahabī says. As for Faḍl, if he is Ibn Marzūq there is some weakness in him; if he is Ibn Ghazwān, he is trustworthy and both al-Bukhārī and Muslim relied on him. Should the latter be the narrator here, the ḥadīth is ḥasan, if Allah wills.
Al-Mundhirī (1/270), after citing the report from al-Ṭabarānī, remarked: “Its transmitters are trustworthy.”
In sum, considering these various routes and supporting reports, the ḥadīth is unquestionably authentic. Ibn Ḥajar’s hesitation about its soundness pertains only to the first chain. Allah knows best.